SCROLL NEWS / DISCOVERY
Search headlines
Find the stories shaping the conversation.
Results for " loses " 867 found 🔀 AI-powered Shuffle
Star Orbiting Black Hole Animation
The precise data collected by NASA’s Hubble and James Webb space telescopes enabled a team of astronomers to chart the visible star’s orbital path over a 20 year-plus period. Researchers recorded this segment of the star’s orbit during its closest approach to it...
Miami Beach Health Alert: Unsafe bacteria closes 3 beaches amid extreme heat
Avondale’s only grocery store closes after a year in business
Rising costs, theft and violence cited as factors in Country Meat Co Marketplace shutdown
'Moana' loses its way at the box office with a $43-million domestic opening
Walt Disney Co.'s "Moana" lost its way at the box office this weekend as the company's latest live-action remake opened to a sluggish $43 million in the U.S. and Canada.
Trump Loses Lindsey Graham
The South Carolina senator’s sudden death leaves a void that the White House can’t afford.
'Moana' loses its way at the box office with a $43-million domestic opening
Disney's live-action remake of 'Moana' came in first at the box office this weekend, but underperformed expectations.
Sen. Lindsey Graham dies at 71. See his life and political career
Sen. Lindsey Graham's death marks the passing of a longtime conservative fixture of the Senate, and one of Trump's closest political allies.
Voyager 1 was launched in 1977 with a suite of 10 science instruments — as of 2026, only two remain switched on, and the spacecraft's nuclear power source loses roughly 4 watts every year, a slow drain that is now deciding what humanity still gets to learn from interstellar space
Voyager 1 is still doing science in interstellar space, but the mission has entered a stage where every watt has become a decision. The spacecraft left
Jacksonville Jaguars QB Trevor Lawrence closes strong at American Century
How did Jacksonville Jaguars quarterback Trevor Lawrence do in the second round of the American Century Championship on July 11?
Microsoft loses Brazilian court case after telling hacked Xbox user to re-purchase games — tech giant ordered to restore Xbox account with all games and pay $400 in damages
A Brazilian gamer who lost his Microsoft account and all his digital games has won a court order requiring the company to return them.
Sports Still Form Character. Politics Should Learn from Them
Athletics may be the closest thing we have to a blueprint for how to disagree well.
Caleb Wilson scores 35 points for Bulls in NBA Summer League debut, but loses 97-96 to Cameron Boozer's Grizzlies
Microsoft loses lawsuit after blocking hacked Xbox account with digital games
One gamer took extreme measures to restore access to his compromised Xbox account. The Brazilian user filed a lawsuit against Microsoft after it refused to unfreeze his digital library. A court ruled in the plaintiff’s favor, and the decline of physical games could lead to more legal challenges.
Mary Steenburgen says ‘Book Club’ cast are her closest friends, they miss Diane Keaton
‘Book Club’ stars Mary Steenburgen, Candice Bergen, and Jane Fonda are extremely close and will be ‘missing’ Diane Keaton when they soon reunite.
Colton's Fiesta Village closes after 50 years in business
Toronto's Dylan Cease loses shot at second career no-hitter in 9th inning of dominant effort vs. Giants
Cease was three outs away from his second career no-hitter.
As Israel Loses Support in the U.S., Rahm Emanuel Criticizes Netanyahu
The former Chicago mayor, a Democrat exploring a 2028 presidential run, is in Tel Aviv calling for an end to unconditional U.S. support of Israel.
Seattle neighborhood loses business during World Cup matches
This Seattle neighborhood lost business during the city's World Cup matches. Here's what community members plan to do before the next big event.
Baltimore Health Department headquarters closes Tuesday after building floods
The Baltimore City Health Department’s headquarters and permits office on East Fayette Street were closed Tuesday after flooding at the building, officials said.
News | Bahraini investor backs South Florida sports and entertainment hub
ME&F Holdings closes first funding round for planned Miami FC stadium
Nature becomes entertainment at award-winning outdoor CT dining spot. Owner calls it ‘oasis.’
The outdoor area, high in demand this time of year and shaded, has three levels: a deck, a patio and the grassy area closest to the river
Report: Arsenal star closing in on £17m summer exit
Leandro Trossard to Besiktas: Arsenal forward closes in on summer exitArsenal are moving towards the sale of Leandro Trossard, with The Athletic reporting that Besiktas have agreed a deal worth “€...
Star-T entertainment reporter, MUTTS closes, why you can't buy a car on Sundays
Catch up with headlines you may have missed this week from the Star-Telegram.
Star-T entertainment reporter, MUTTS closes, why you can't buy a car on Sundays
Catch up with headlines you may have missed this week from the Star-Telegram.
Valley Golf Center owners say it's been an 'honor' as business closes
KMPH Fox 26 is the Central Valley's news leader, covering Fresno, California and the surrounding area, including Clovis, Madera, Hanford, Visalia, Biola, Kerman, Tranquillity, San Joaquin, Reedley, Coalinga, Avenal, Lemoore, Tulare, Selma, Caruthers, Tollhouse, Three Rivers, Porterville, and Merced.
Ecolab Closes CoolIT Acquisition and Expands AI Cooling Platform as Global High Tech Business Targets $4 Billion by 2030
ST. PAUL, Minn., July 02, 2026--Ecolab today announced the close of its acquisition of CoolIT Systems, a leader in direct liquid cooling for high‑density data centers. The transaction closed earlier than expected, and Ecolab welcomes the world-class CoolIT team to the company.
From The Sports Desk: The U.S. advances, but loses Folarin Balogun to costly red card
The U.S. beat Bosnia-Herzegovina 2-0 to advance to the round of 16, where it will face Belgium next. But the U.S. will be without Balogun, its top goal-scorer of this tournament.
Controversial penalty ends Senegal’s FIFA World Cup run against Belgium
Late VAR penalty leads to fury as Senegal blows 2-0 lead, loses 3-2 to Belgium in dramatic World Cup Round of 32 exit.
Sony will cease production of discs for new PlayStation games in 2028
New games sold for PlayStation consoles will only be available digitally and won’t be sold on discs effective January 2028, Sony has announced. The move applies to all PlayStation games released from that point on, including those made by Sony and those released by third parties. The PlayStation group calls the change “a natural direction for Sony Interactive Entertainment to adapt to consumer trends as the general preference for digital media significantly outpaces physical discs,” according to a PlayStation blog post reviewed by Game File in advance of its publication. As movies, TV and music have shifted from physical media to an age of downloads and streaming, Sony’s move could speed the demise of physical media in a gaming. The industry has offered consumers games via cartridges, cassettes, floppy discs, CDs, DVDs and Blu-Rays across a 50-year stretch. Sony has been reporting to investors in recent years that PlayStation games have been increasingly purchased as downloads, not as discs. The most recent figures show nearly four in five purchases of full games for PlayStation 4 and PlayStation 5 were purchased digitally in the past year. Stats like that may cause today’s news to carry an air of the inevitable. But Sony’s declaration for 2028 is still likely to produce a shock, as the biggest gaming platform on the planet that uses game discs leaves them behind. Discs going away, but the fate of game boxes is unknown Sony is committing to still selling games in physical retailers even after they drop discs in 2028. Just how those games will be sold—in boxes with codes inside, as cards marked with digital redemption codes—is unclear. Digital games will also continue to be sold in the online PlayStation Store. New games released for PlayStation consoles prior to January 2028 will be unaffected by the policy change, Sony says. That means the PS5 games of 2027, be they a new Madden, a new God of War, a new Call of Duty, a new Sly Cooper (one can dream?), would still potentially be offered on disc. PlayStation 6 implications Sony’s timeline for a transition to a no-disc era will also spin up more speculation about the timing and configuration of Sony’s next-generation console, the probable PlayStation 6. “This pretty much guarantees that PS6 won’t arrive until 2028 at the earliest,” Piers Harding-Rolls, senior games research analyst at Ampere Analysis tells Game File. Harding-Rolls also concludes from today’s news that “the base version of a PS6 will not include a physical media drive,” as Sony looks to keep costs down on the device. Sony itself is saying nothing about the PS6 today: not its timing, nor how it might or might not support internal or external disc drives in order to run disc-based PS4 and PS5 games, should the console be backwards compatible. The many signs this was coming Industry support for games being released on physical media has been waning. As has consumer interest in them. PC games effectively went all-digital years ago, thanks to the popularity of playing games through Steam and other digital storefronts. The Covid-19 pandemic and the lockdowns that kept much of the global population home helped shift shopping patterns toward downloads. Sony’s own data shows PlayStation game downloads being an uncommon consumer choice a decade ago and now becoming the norm. On consoles, physical media has been in such decline that Sony and Microsoft both launched their current gen Xbox Series and PS5 consoles in 2020 with cheaper versions that lacked disc drives. In early 2024, Phil Spencer, then head of Xbox, told me that ““[w]e ship games physically and digitally, and we’re really just following what the customers are doing.” He added that the game industry had become an outlier in still supporting disc-based entertainment, even as movies and TV have bailed on discs in favor of downloads and steaming: “Gaming consoles themselves have kind of become the last consumer electronic device that has a drive. And this is a real issue, just in terms of the number of manufacturers that are actually building drives and the cost associated with those. And when you think about cogs that we’re going to go put in a console—and as you have fewer suppliers and fewer buyers—the cost of the drive does have an impact.” Last year, the Nintendo’s Switch 2, which uses cartridges rather than discs, introduced the concept of game-key cards, which are physical media that contain no game, just a prompt for a download. Even disc-based games aren’t what they once were, now frequently containing just stubs of games rather than the whole thing. The website Does It Play has tested over 3000 games and found that, for example, that the disc version of this year’s hit 007: First Light, only contained the game’s first mission, with the rest needing to be downloaded. Beyond all that, last week brought the news that Grand Theft Auto VI, likely the biggest game launch this decade, will debut for PS5 and Xbox Series without a disc-based version. Today’s big news comes to you for free. Game File is a reader-supported publication. To receive new posts and ensure more of my work is possible, please consider becoming a free or paid subscriber. Subscribe Discs helped define PlayStation However the trends in entertainment may be going, it is still remarkable to see PlayStation ditch discs given their importance in PlayStation history: In the late 1990s, the first PlayStation’s support for games on relatively cheap CD-ROMs instead of traditional game cartridges allowed the upstart console to outmaneuver cartridge-based systems from Nintendo and Sega, as it drove down game prices down in the process. (The cheaper disc format even prompted Final Fantasy VII maker Square to taunt Nintendo in a magazine ad, saying that if the game was “available on cartridge, it’d retail for around $1200.”) PlayStation devices often doubled as machines for watching movies and TV on the latest disc-based formats. In the late 2000s, for example, Sony’s pricy PS3 gained traction as a go-to device for playing then-novel Blu-Ray discs. In 2013, Sony effectively roasted the digital game-sharing restrictions proposed by rival Microsoft for its upcoming Xbox One console by showing how games could be shared for the then-upcoming PS4. In a 21-second clip that has over 20 million views on YouTube, one PlayStation executive simply handed a game box—presumably with disc inside—to another, who said “thanks.” Sony’s possible motivations The PlayStation announcement today cites “consumer preferences” and a shift away from discs among “the broader entertainment industry,” but does not delve into the decision-making process that brought Sony to this point. Sony is not discussing the economics of this move (I asked!), but there are a number of potential factors: Abandoning discs will spare Sony the cost of manufacturing and shipping them, which analysts and industry veterans I spoke to said could account for $1 or so of a game’s retail price. For first-party games, that’s direct savings for Sony, though the amount is limited by the winnowing percentage of people who were even buying the disc-based versions. For third-party games, Sony could pass the savings on to partners or just retain the better profit margin for themselves. As the 2024 comments from Phil Spencer alluded to, leaving discs behind could mean leaving disc-drives behind, allowing Sony to produce consoles more cheaply. Switching to digital-only for games could diminish the risks of copies of new games leaking. It would also seemingly make second-hand sales of used games impossible, pushing more people to buy their own copies of games. These economic benefits for Sony could be boons for the games industry overall, Ampere analyst Piers Harding-Rolls tells Game File. Among other things, it could free publishers and retailers from having to worry about securing an inventory of game discs when trying to sell games, he noted. “If there is a way to streamline the business of games retail, I generally think it’s a good thing for the industry as it removes unnecessary costs,” Harding-Rolls said. “This will help offset pressure on margins from other directions – development and staffing costs for example.” Should the industry shift from game boxes that contain discs to smaller, cheaper game cards, Harding-Rolls said, that reduces pressure on publishers and retailers. “The advantage of everything going digital for the industry is that sales can be tracked much more easily and there is less commercial risk for publishers wanting to continue selling via retail,” he said. The shift to no discs will also likely impact game development timelines, as developers will no longer face the same urgency to have a build ready to submit to PlayStation for the production of a disc, as industry experts noted around GTA VI’s no-disc news last week. Instead, projects may gain weeks of development time, in the event delivery deadlines for platform approval move closer to release date. Preservationists’ likely worries While millions of gamers have shifted to digital, there are still people who associate game discs—and only game discs—with actual ownership of PlayStation games. That portion of the gaming community spoke up last week about wanting to buy a version of GTA VI that they could hold in their hands, could lend to a friend and could potentially pop into a well-preserved PS5 console three decades from now and still see it run. Such gamers didn’t want to worry that they’d merely purchased a digital copy that could lose access to the game if, say, the company behind the game someday turns of the servers that transmit downloadable copies. Digital copies of games are inherently less long-lasting, because consumers aren’t actually buying a game (on a disc), just buying access to the game. The fine print in online gaming shops notes this, as does a recent law that went on the books in California in 2025. That lasw requires game publishers doing business in the state to refrain from words like “sell” when offering digital copies of games, unless they explicitly tell consumers that what’s being sold isn’t the game itself but a potentially revocable license to access it. A Sony spokesperson noted to Game File that “[w]ith all digital content, including games, movies, and music, players are purchasing a personal license for non-commercial use.” Regarding the fears of PlayStation games being lost to time in an all-digital era, the rep also noted that Sony remains “committed to delivering experiences from past generations to the new platforms our players are using.” They noted that PlayStation has used emulation and remasters to keep offering its classic games to modern gamers. It’s true that many consumers have become accustomed to not having physical copies of new music albums or movies, but the relative precariousness of digital copies of games is in fact being highlighted today by…. Sony’s other big announcement: the upcoming closure of the PS3 and PS Vita online stores Sony said today that it will be closing the online marketplaces for its PlayStation 3 console and PS Vita handheld in most countries in July 2027 (sooner elsewhere: August 2026 for the PS3 store Mexico, Honduras and Nicaragua; late 2026 for PS3 in other Latin American countries and the Middle East). “We know this news may be disappointing to PS3 and PS Vita players who hold a special place in their hearts for this generation of gaming,” Sony said in its announcement, saying the move “was not an easy decision for us to make.” Sony said that PS3 and Vita marketplaces “are no longer able” to support modern commerce systems, prompting a “need” to close the shops and focus on more modern PlayStation platform stores. The news may spark some deja vu. Sony announced back in March 2021 that it planned to take the PS3, Vita and PlayStation Portable marketplaces offline that summer. The move would have caused some 2,000 digital-only games to become unavailable for new purchasers, according to a VGC analysis at the time. Fans complained loudly. By April 2021, Sony walked back the PS3 and Vita shop closures, with then PlayStation boss Jim Ryan saying “it’s clear that we made the wrong decision here.” This time around, Sony is giving a longer notice period in most regions. The closure of those stores is still likely to cut off access to new purchases of digital-only PS3 and Vita games. The current PlayStation 5 marketplace sells PS5 and PS4 games, as well as a selection of PS1 and PS2 games. It does not sell PS3 games, because games for that platform can’t run on the system (some are offered via a PlayStation Plus streaming option). Vita games also aren’t sold via PS5. Sony says digitally purchased PS3 and Vita games will be downloadable after the stores’ shutdowns “for the foreseeable future.” Sony’s move contrasts with Microsoft’s Xbox platform, which continues to sell and support games from all four Xbox console generations. But it puts it in line with Nintendo, which disabled purchases in the marketplace for its Wii console (a contemporary of the PS3) in 2019. Nintendo discontinued purchases in the online shop for its 3DS portable (a contemporary of the Vita) in 2023. The big questions about Sony’s 2028 post-disc era Will Microsoft follow suit, ending support for new disc-based games on Xbox platforms and therefore all game consoles, too? Whether Microsoft does or not, will third-parties that can no longer release games on discs for PlayStation, simply ditch the disc options across the board? Given the GTA VI no-disc experiment launching this November, will more game publishers decide to forgo disc-based offerings even in 2027? How will no-disc PS5/6 games be packaged and sold in brick-and-mortar retail stores? What becomes of game sharing in an all-digital era? Will Sony modify its current set-up, which currently allows some sharing of digital games across a pair of consoles, to allow more casual lending? (See for example, Nintendo’s year-old system of letting players lend a digital game to another player in a “family group” for up to 14 days.) Will there be a version of a PS6 with a disc drive? Or will a drive be offered as an add-on? And how will gamers who want to retain access to PS4/PS5 disc-based games react to whichever configurations Sony chooses to offer? What will the final disc-based PlayStation game be?
Serena Williams loses in opening round at Wimbledon
Williams, who is 44 years old, is a 23-time Grand Slam tennis champion
Speaker Johnson loses key vote as House GOP leaders struggle to contain rebellion
Speaker Mike Johnson is once again battling hardline conservatives for control of his own House floor — and he has no clear way out.
NATO joins exercises off US coast, even as it loses political fight in Trump's Washington
The warning came booming over the airwaves, in a voice unmistakably from the American south: "Warship! Get out of our waters!"
'You're so full of it': Gee blasts Seattle's 'performative' homelessness politics as city loses nearly 700 shelter beds
Seattle's homelessness policies fall short, with nearly 700 shelter beds cut, revealing a need for more effective solutions.
Republic bakery owner loses home and business in weekend fire
A bakery and boutique that had operated in Republic for nearly three decades was destroyed by fire over the weekend, leaving its owner without her business or her home.
Opinion | America Is the Land of Endless Choice, Except in Politics
There’s a case to be made for a two-party system, but what do you do when your party loses its mind?
Kidambi Srikanth loses US Open final: Indian Sports LIVE, June 29
Here are all the updates from Indian sports on June 29, 2026.
Beloved Brookline restaurant closes after over 60 years in business
A well-known Brookline restaurant announced on Instagram that it officially closed for good after over 60 years in business.
After decades in the business, a Staten Island owner closes his dream restaurant
After decades building his dream in Annadale—and nearly losing it during COVID—the restaurateur steps back to savor family milestones.
Colonial Market owner closes south Minneapolis location due to sharp drop in business following Metro Surge
Colonial Market owner Daniel Hernandez has closed the chain’s south Minneapolis location, which he said never recovered after the store’s employees and customers were targeted during Operation Metro Surge.
'A heavy impact:' Schwebel's closure leaves a sense of community loss
It’s the small things that residents across the region will remember when Schwebel’s closes its doors at summer’s end. Everything from...
Need for bread | Unexpected closure of Terranetti's Bakery has left several businesses scrambling
Terranetti's Bakery closes, and local shops are scrambling for bread
Cheshire orchard closes after 75 years in business
A longtime Cheshire orchard has announced its closing for good. In a message shared with customers, the family that owns Drazen Orchards said it is now permanently closed and will not reopen for the foreseeable future. The orchard has been open since 1951. The Drazen family said the decision was difficult, citing rising operating costs, the physical demands of running…
Hollywood production house closes, citing entertainment industry struggles
Hollywood production house closes, citing entertainment industry struggles
South Korea juggles digital sovereignty and alliance politics
South Korea’s closest allies are increasingly challenging its control over digital platforms, data and technological infrastructure.
Kennedy scion Jack Schlossberg loses crowded New York City congressional primary
Kennedy family scion and political novice Jack Schlossberg lost Tuesday to New York state Assembly Member Micah Lasher, in a closely watched and crowded Democratic primary for an open congressional seat in the heart of Manhattan.
Hancock eatery closes after eight months, owner puts business up for sale
After just eight months in business, Roman’s Bistro & Bar is closed and the business is for sale for $62,000.
Goldman loses House seat after NY-10 Democratic primary defeat to Lander
NBC News projects that Lander, a progressive backed by Mayor Zohran Mamdani, won the Democratic nomination in the 10th District.
Ferrara’s Imported Foods in Mayfield Heights closes after nearly 70 years in business
After more than 65 years in business, Ferrara’s Imported Foods in Mayfield Heights has closed its doors for good.
South Portland dog park/bar closes after 4 months of business
A Maine restaurant and bar that included an indoor dog park has closed less than five months after it opened for business.
Romania’s Political Turmoil Deepens After PM Hopeful Loses Vote
Romania’s prime minister nominee failed to win parliamentary approval for his proposed cabinet, prolonging a political deadlock that risks stalling key reforms for crucial European Union recovery funding.
Putnam Science Academy closes after losing accreditation, facing financial turmoil
Putnam Science Academy, a for-profit boarding school known for its standout basketball program, has closed, according to a letter attributed to school president Tieqiang Ding that was circulating on social […]
SpaceX launches notes offering, discloses $100.8 billion cash balance
SpaceX said on Monday it launched an offering of senior unsecured notes and disclosed that it held about $100.8 billion in cash and cash equivalents as of June 19.
Health clinic in St. Louis loses federal funding. CareSTL was already collapsing.
CareSTL Health, which serves low-income and uninsured patients, has already shuttered four of its five locations.
A 'perfect storm’: Why rural hospitals across Michigan are struggling
As Sturgis Hospital closes, there have been concerns raised over healthcare access in southwest Michigan.
$4M sale closes for KC Wheel site once pitched as larger entertainment district
KC Wheel site changes hands in $4M sale.
Trump wins two, loses one: Georgia billionaire delivers rare blow to endorsement machine
Trump-backed candidates won two of three key GOP runoffs in Georgia and Alabama, but lost the gubernatorial race to billionaire Rick Jackson.
Montana Board of Investments closes deal on Laurel land for proposed forensic mental health facility
The state has closed on its purchase of 114 acres in west Laurel proposed for a forensic mental health facility and is now asking Yellowstone County to rezone the property.
Board recommends sheriff changes, 40-year business closes among top TNT stories
Also Tuesday, the latest restaurant inspections and The News Tribune’s food writer offers her thoughts on Knapp’s demise.
Kylian Mbappé passes Lionel Messi on all-time World Cup goals list, closes in on Miroslav Klose's record
Xbox Game Pass loses another eight games in June 2026
The next wave of games leaving Xbox Game Pass in June 2026 has been confirmed. In total, eight games will be removed from Game Pass at the end of the month.
Taj Clubhouse closing at Ward Centre after 50 years in business
Shoppers grabbed what they could before the family-owned store, known for its hand-drawn, locally inspired characters, closes for good.
Designing the Health Systems AI Promised
Every technology transition in healthcare has produced the same pattern: capability arrives before governance, deployment outpaces integration, and the accountability gap closes, eventually, through incident, litigation, or regulation. AI is running the same pattern, faster and at greater scale, in systems that were already fragile before it arrived.The more uncomfortable question shifts the lens away from what AI is doing to health systems, toward what health systems were never built to be. The institutions now absorbing...
USC Baseball Loses Another Player to Transfer Portal
USC already lost star hitter Augie Lopez after he entered transfer portal, and now they're set to lose a pitcher as well. The post USC Baseball Loses Another Player to Transfer Portal appeared first on LA Sports Report.
Topeka Renaissance Festival closes out weekend of entertainment
TOPEKA (KSNT) – The Topeka Renaissance Festival wrapped up its three-day run Sunday, bringing together performers, vendors and visitors for a weekend of costumes, live entertainment and historical-…
Business roundup: Garden center closes, cookware shop opens and Ore House timeline uncertain
In this week’s business roundup: Botanical Concepts Garden Center in northeast Durango decided to call it quits this spring, a new cookware shop is set to open on Main Avenue, and a timeline for the r...
Is SMIC N+3’s Metal Pitch Smaller than Intel 18A’s?
Almost four years ago, we published that SMIC had started shipping 7 nm (N+1) chips. Now, SMIC is shipping its third-generation 7 nm (N+3) process in Huawei’s Kirin 9030, with a minimum metal pitch of 32.5 nm, about 10% tighter than the 36 nm minimum metal pitch shipping in Intel’s latest Panther Lake CPUs on 18A. The headline is true, but incomplete cherry picked metric. N+3 reaches the density of TSMC N6 through aggressive DUV multi-patterning and design-technology co-optimization (DTCO), but it pays for that in complexity, efficiency and process control. We found this and more in our reverse engineering and teardown where we cover SMIC’s N3 process technology, Huawei’s packaging, memory, architecture, and more. SemiAnalysis has been building a state-of-the-art teardown lab in Oregon capable of analyzing the world’s most advanced and important chips over the last year and half. We have already generated revenue on advanced datacenter chip teardowns including our recent reverse engineering of a major TSMC customer’s COUPE CPO optical engine + EIC 3D stack. This is the first public report from the SemiAnalysis Teardown Engineering & Evaluation Lab, or STEEL for short. The lab is aggressively scaling up and out and we’re excited to announce it publicly. This is a bit of inconvenient timing for TechInsights as they are private equity owned and currently being sold while having enjoyed virtually no credible competition for decades. This has led to TechInsights underinvesting in CAPEX. SemiAnalysis exceeds TechInsights in revenue despite no venture or private equity ownership and being founded only 6 years ago. Because we have no external investors and are founder led, we move faster, build faster, and we can release client chip teardowns for free regularly, while focusing on datacenter for our major clients. Here’s the first public image from our lab, the HiSilicon Kirin 9030 Pro SoC: This report will detail our teardown of the Kirin 9030 and our findings on SMIC’s N+3 process, the most advanced in China. For comparison, we’ll show our teardown of the MediaTek Helio G99, made on TSMC N6. Through this comparison, we can look at the effect of export controls – SMIC N+3 and TSMC N6 are comparable nodes, but one is heavily export-controlled, the other free to use the West’s most advanced equipment. Here we see both China’s progress and constraints. SMIC N+3 reaches TSMC N6-class logic density, but it requires far more aggressive DUV multi-patterning, so it does not match N6 on process maturity or cost. The Kirin 9030 Pro performs similarly to three-year-old Android flagships, and trails far behind the current flagship SoCs from Apple, Qualcomm, MediaTek, and Samsung. The efficiency gap is even wider. Export controls have not stopped Huawei and SMIC from shipping advanced silicon, but they have forced a different path. Without EUV, SMIC is leaning harder on DUV multi-patterning, DTCO, and increasingly complex integration. The roadmap continues forward through tighter design rules and backside power, but each step adds cost and process risk. Huawei’s τ scaling and LogicFolding show another path: stacking active logic and recovering density through advanced packaging and system-technology co-optimization (STCO). To understand the Kirin 9030, we must first understand Huawei’s SoC history. HiSilicon is Huawei’s chip design arm, responsible for the Kirin smartphone SoCs, Kunpeng server CPUs, Ascend AI accelerators, and switch/router networking silicon. Before export controls, Huawei was TSMC’s largest customer – the only customer on TSMC’s first EUV node, N7+, and among the first on N5, alongside Apple. That ended in late 2020. Huawei switched to Qualcomm SoCs in its flagship smartphones, though export controls limited them to 4G-only variants. In late 2023, Huawei returned to in-house silicon with the Kirin 9000s, a successor to the Kirin 9000, fabricated on SMIC N+2 instead of TSMC N5. In the following years, they released the Kirin 9010 and 9020 on the same N+2 process. These chips used Huawei’s in-house TaiShan CPU cores and Maleoon GPU. We have not torn down a Kirin 9020 ourselves, so the predecessor die shot is from Kurnal. The die shots show how Huawei spent its silicon budget: which functional blocks are where, and how their areas compare to the predecessor. First, a quick guide to the major blocks on the die. The total die area is nearly identical, but the 9030 uses that area more aggressively. A denser process lets Huawei fit an extra middle CPU core, more GPU and NPU cores, and larger caches into the same footprint. In contrast, the Helio G99 is a much smaller, low-cost SoC, built for budget smartphones rather than a flagship device. While the Kirin 9030 is ~140 mm², the G99 is only ~29 mm², roughly one-fifth the area. The underlying TSMC process technology, however, is directly comparable as a baseline for analyzing SMIC’s. The Kirin 9030 is an evolutionary refresh, not a clean-sheet design. Its CPU, GPU and NPU cores carry over the 9020’s families, and the gains come from three levers: the SMIC N+2-to-N+3 process step, DTCO and floorplan work, and incremental microarchitecture. Area is where the first two show up, and the 9030 scales well here. Performance and efficiency are the harder test. Huawei’s design holds up better than its node would suggest, but the chip still trails, both because N+3 sits behind the leading-edge nodes and because its cores, while competent, remain a few generations behind the newest designs. The new prime core is an incremental update. The main changes are a 10% frequency increase from 2.5 GHz to 2.75 GHz and a doubling in the L2 cache from 1 MiB to 2 MiB. Despite the increased cache, the core size decreased by 7.6%. Excluding the private L2 cache, the core size decreased by 21%. This is a large reduction for an incremental node. Compared with the TaiShan New V120 core in the Kirin 9020, the Kirin 9030’s middle core is almost unchanged architecturally, yet each core shrinks by ~22%. Most of that comes from the move from N+2 to N+3, with layout likely accounting for the rest. Visually, the most noticeable change is the increase from 3 middle cores to 4. There is also a 20% increase in the shared L3 cache of the big cluster. This helps to improve multi-core performance without sacrificing much in terms of area. Even with each core shrinking, the big CPU cluster’s total area is essentially unchanged. The per-core savings went back into an additional middle core and larger caches. The tiny cores shrank less than the prime core (excluding its L2 cache) and less than the middle cores. This is likely because fixed overhead is a larger share on a small core. We cannot resolve any architectural changes from the die shot alone, but the per-clock and efficiency gains shown below point to more than pure process and layout scaling. The area reduction was offset by a doubling of the shared L2 cache from 2 MiB to 4 MiB, leaving the total tiny CPU cluster area slightly larger. Area is the easiest improvement to see from a die shot, but it is only one part of PPA (power, performance, area). For modern logic, power and performance matter just as much, and often more. Since Dennard scaling broke down in the mid-2000s, voltage and frequency have not scaled in step with transistor dimensions, so each node has had to fight harder for gains in performance and efficiency. The starkest comparison is not Kirin 9020 versus Kirin 9030 Pro. Apple’s efficiency cores run circles around Huawei’s prime core. Apple’s low-power core delivers 20% higher integer performance while drawing only 1 W, compared with 4.5 W for Huawei’s prime core. N+3 matches TSMC N6, but N6 is several generations old. Apple and Qualcomm build on N4 and N3P, which are denser and sit on a better voltage-frequency curve, giving them a larger transistor budget and more performance per watt. The 9030’s own cores did improve. The middle and tiny cores gained 17% and 14% in per-clock integer performance over the 9020, with floating-point flat on the middle core and up 11% on the tiny. The tiny core improves cleanly, with performance rising while power falls and efficiency increases by 45% in integer and 24% in floating point. The middle core is mixed: integer performance rises but power rises faster, cutting integer efficiency by 7%, while lower power lifts floating-point efficiency 16%. Per-clock gains at the same or lower frequency are microarchitectural, so the cores are tuned, not just shrunk. Both also failed to hold their rated maximum frequencies, pointing to thermal, power, or stability limits. Per clock, the middle core sits around Arm Cortex-A720 and the tiny core near the Cortex-A520; absolute performance trails because Huawei clocks them much lower. The prime core is roughly Cortex-X2 class per clock, a 2021 design. Apple’s 2020 M1 Firestorm core is still 35% higher per clock and 57% faster in absolute integer performance at a similar 4.5 W. The current leading edge is further ahead again: the Apple M5 P-core is 60% higher per clock and 2.7× faster, the Arm C1 Ultra 45% higher and 2× faster. Matching older high-end cores per clock is a genuine design achievement. What Huawei cannot match is the voltage-frequency curve and transistor budget of leading-edge nodes, which let Apple, Qualcomm and others spend more transistors in the same area on wider cores, larger caches and deeper buffers while running at lower voltage. Huawei’s LogicFolding roadmap is one answer, stacking active logic to recover density and shorten signal paths. We return to it later. The GPU compute units (CUs) changed more visibly than the CPU cores, moving to a more rectangular layout for both the arithmetic logic unit (ALU) clusters and the CU overall. Even with ray-tracing support added, a CU shrank ~28%. However, that shrink is offset by the increase from 4 to 6 CUs and the area outside the CUs grew 33%. Overall, the GPU cluster is larger by ~10%. The GPU is where Huawei makes its biggest gains. The Maleoon 935 is not competitive with current flagships, but it is a large step up from the 920 and reaches older-flagship territory. In 3DMark it is 70% faster in Wild Life Extreme (WLE) and 79% faster in Steel Nomad Light (SNL) than the 920; with 11% higher clocks and 50% more CUs, the ~67% theoretical uplift roughly matches WLE and is beaten by SNL. It edges ahead of the Snapdragon 8+ Gen 1 in WLE and SNL, and the Dimensity 9200 and Apple A16 in WLE, but stays far behind newer parts: the Snapdragon 8 Elite Gen 5 and Dimensity 9500 are ~2.4–2.6× faster in WLE and ~3.2× faster in SNL. The Maleoon 935 is Huawei’s first GPU with hardware-accelerated ray tracing; there it lands slightly ahead of the Exynos 2200, and on par with the Apple A16, with current flagships up to 3.7× faster. The Neural Processing Unit (NPU) saw the largest structural changes of any block, moving from a Lite and a Tiny core in the Kirin 9020 to a Lite and two Tiny cores in the Kirin 9030. Both core types also show significant layout changes. This is a reversal in Huawei’s NPU design. The Kirin 9000 5G, its last flagship chip on TSMC N5, used two Lite and one Tiny core. The series of SoCs on SMIC N+2 moved to one Lite and one Tiny core, likely for area savings. With the Kirin 9030, Huawei has shifted back toward a larger multi-core NPU cluster, but with the additional area going to a Tiny core rather than a Lite core. We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline or to commission a custom teardown, contact sales@semianalysis.com. Interested in joining us on this ride and think you can be a difference maker? Check out our Careers page. Before diving into the process stack, the package and memory are worth separating from the SoC itself. The Pro variant of the Kirin 9030 carries 12 GB of Samsung DRAM, with two stacks of four dies each. The dies were identified as the K4L2E165YD, a 12 Gb LPDDR5X-9600 device fabricated on Samsung’s 1a node, the fourth generation of its 10 nm-class DRAM after 1x, 1y and 1z. 1a has shipped in volume since 2022, so this is current memory rather than older-node inventory. The 16 GB Pro Max variants we obtained were found with both CXMT and Samsung packages. The CXMT package is marked CXDD7JEDM, with two stacks of four dies, packaged in week 45 of 2025. The inferred die dimensions from X-ray computed tomography (CT) are consistent with a known density of ~0.3 Gib/mm² for the CXMT G4 process, roughly equivalent to other manufacturers’ 1z processes. The Kirin 9030 uses a typical integrated package-on-package (iPoP) stack: multiple DRAM dies in a memory package sit above an organic redistribution layer (RDL) interposer, which sits above the SoC and package substrate. The full package is then mounted to the printed circuit board (PCB) through ball-grid array (BGA) solder bumps. The memory package substrate is a thin bismaleimide-triazine (BT) laminate carrying the LPDDR5X stack. The organic RDL interposer over the SoC routes the PoP signals around the die and carries possible dummy thermal copper pillars. The package substrate, a thicker Ajinomoto Build-up Film (ABF) build-up over a BT core, fans the flip-chip bumps out to BGA pitch and embeds the power planes. The whole stack is organic. The only silicon is the SoC and the LPDDR5X dies; there is no silicon interposer. Keeping it all-organic brings the package’s coefficient of thermal expansion (CTE) close to the PCB’s, reducing board-level warpage, and avoids the cost of a silicon interposer the SoC’s bandwidth does not need. In an iPoP stack, the memory package connects to the organic RDL interposer through an array of solder bumps. Underfill fills the gap around those bumps, adding stiffness and protecting the joints from mechanical stress. The Pro and Pro Max variants differ here, which we cover behind the paywall. The die shot and architecture tell us how Huawei allocated its silicon budget. The process tells us what SMIC can manufacture. We use the Helio G99 as the process reference for TSMC N6. Both SMIC N+3 and TSMC N6 are evolutions of previous 7 nm-class nodes. We used targeted TEM cross-sections through logic and memory regions, imaged in both fin-cut and gate-cut directions. Each cross-section caption gives its horizontal field width (HFW), the real width of the imaged area. We start at the transistor fins, then move up through standard cells, local interconnect, and SRAM. SMIC has not overtaken Intel or TSMC. It uses aggressive DUV scaling and DTCO to reach N6-class density, but that density doesn’t translate into comparable performance and efficiency, for two reasons: the node gap to leading-edge nodes, and Huawei’s core designs. Fin Profile One of the most important knobs in a FinFET process is the fin profile: the shape of an individual fin and the channel where current passes from source to drain. The ideal fin is tall, narrow, and nearly vertical. A taller fin increases effective channel width, while a narrower fin improves electrostatic control by thinning the body the gate must control. Push either too far, and the process pays for it: weaker drive current, fragile fins, taper, footing and line-edge variation that hit yield and device variability. The Intel 22 nm, 14 nm, and 10 nm fin cross-sections show how FinFET nodes have improved over time. 22 nm fins were a first-generation structure, relatively short, wide and strongly tapered. The shape limits current density and reduces gate control uniformity across the height of the fin. At 14 nm and 10 nm, Intel pushed the fins taller and narrower while also making the sidewalls more vertical. Rather than shrinking the device, these changes increase the effective channel width per fin and improve electrostatic control. The trade-off is that taller fins at tighter pitches are much more difficult to manufacture. Now, let’s compare the Helio G99 on TSMC N6 with the Kirin 9030 on SMIC N+3. Both processes are in the same class, with fin pitch of 30-32 nm on N+3 and 34 nm in our N6 cross-section. The pitch for N6 is especially interesting as N7’s HD library is generally listed with a 33 nm fin pitch, and N6 did not shrink pitches directly. Its density gains came from DTCO instead of tighter pitches. The 34 nm pitch was stable across our sampled region and serves more as a comparison against the SMIC N+3 we have not investigated further. Pinning down N+3’s fin patterning scheme takes more than one core unit. The CPU cores show a dense ~32 nm pitch, with the pitch between N-P fin pairs alternating between 78 and 88 nm. Logic alone may be consistent with dual-pitch mandrels of 120 and 110 nm, but this is a complex and unusual approach. Combining the pitch from the the 8T SRAM, which has more complex repeat unit, with the CPU core sequence allows us to reverse engineer the patterning steps with more confidence. As both the logic and SRAM should share the same base grid, a single CD mandrel lithography pattern with 128 nm pitch undergoing SAQP produces a die-wide ~32 nm grid (128 nm/4), which supports the pitch sequencing seen in both logic and SRAM cells. In the sampled cross-sections, N+3 shows a taller, narrower, higher-aspect-ratio fin than N6. The measured fin aspect ratio is ~9.5:1 on N+3 versus 7.8:1 on N6. N+3 also shows less top rounding, with an estimated radius of ~2 nm, compared with 2.8 nm on N6. Even though the fin widths differ, the ratio of top rounding to fin width tells the same story, with N+3 at 0.37 and N6 at 0.44. In a geometric sense, lower is better; a perfectly rectangular fin would have no top-rounding penalty. These are single-digit-nanometer features measured from a handful of cuts, so treat the absolute numbers as approximate. The important result is the relative gap: N+3’s fins are consistently taller, narrower and less rounded than N6’s. We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline or to commission a custom teardown, contact sales@semianalysis.com. Interested in joining us on this ride and think you can be a difference maker? Check out our Careers page. Standard Cell A standard cell is the basic building block of chip layout: a fixed-height row pairing one NMOS and one PMOS transistor that share a gate, tiled in a grid to build logic blocks. The key dimensions are contacted gate pitch (CGP), cell height (CH), fin count, and the lower-metal routing grid. To measure density, we use the Bohr metric: a weighted average of NAND2 gate area (60%) and scan flip-flop area (40%). This represents a realistic mix of combinational and sequential logic. This metric has its limitations, especially for complex cell layouts like TSMC’s FinFLEX, which alternates cells with different fin counts. Even so, it is the best metric for a pure process-level comparison. Another important measurement is the fin pitch; it refers to the distance between two fins of the same transistor. In a FinFET process, multiple fins are used in each transistor to increase the drive current and thus performance. TSMC N6 ships both a high-density (HD) library with 2 PMOS and 2 NMOS fins per cell, and a high-performance (HP) library with 3 of each. More fins under the shared gate mean more effective channel width. HP cells switch harder at the cost of area. Designers mix the two on a die, primarily spending HP cells on timing-critical paths, and matching their PPA targets. In the Cortex-A55 core of the Helio G99, we found a cell height of 240 nm for the HD cell. MediaTek has used HD cells in the G99 to minimize die size and thus cost. As an SoC for budget smartphones at ~$100, this is essential. By contrast, we found only one library in the Kirin 9030, with 2 NMOS and 2 PMOS fins. This suggests a narrower library strategy than TSMC N6, where both HD and HP libraries are widely used. This likely reflects the smaller customer base and the more constrained domestic design and electronic design automation (EDA) ecosystem. In all three CPU cores of the Kirin 9030, we found cell heights of 228 nm, 5% smaller than on N6. This is also a reduction of 9.5% over SMIC N+2’s cell height of 252 nm. SMIC N+3 and TSMC N6’s HD library both feature a CGP of 57 nm. For SMIC, this is a 9.5% shrink over N+2. In the past, CGP and cell height alone may have been enough to compare transistor density. Now, however, we must consider scaling boosters and DTCO as well. SMIC’s density gain does not come from EUV. It comes from using every available DTCO booster aggressively. First is fin depopulation: reducing the number of NMOS and PMOS fins in each cell. The first FinFET nodes started with 3 or 4 fins for each transistor. SMIC N+3 and TSMC N6 HD both use only 2 fins per transistor, trading drive strength for density. Next is contact over active gate (COAG). By landing the gate contact directly over the active gate, instead of out over the isolation region, the cell height drops. N+3 integrates COAG while N6 does not. Our N+3 gate-cut cross-sections indicate COAG, with the gate contact sitting over the active region, while N6 shows an off-gate contact. Last is single diffusion break (SDB). Diffusion breaks are inserted between cells in the same row to provide electrical isolation, but they also introduce local layout effects (LLE), layout-dependent shifts in electrical characteristics. In the past, a double diffusion break was used, consuming the space of two CGPs. SMIC N+3 and TSMC N6 instead use SDB, saving area but increasing LLE sensitivity. This must be controlled at the process level and accurately modeled in the process design kit (PDK) so that EDA tools can account for it. Overall, SMIC N+3 has a transistor density of 113.4 MTr/mm², slightly above TSMC N6 at 107.7 MTr/mm². Even without EUV, SMIC has achieved density beyond TSMC’s mature N6 node which utilizes EUV. Metal Stack The smallest critical dimension in the teardown is M0; SMIC N+3 uses a 32.5 nm local metal pitch. That is smaller than the 36 nm M0 pitch on Intel 18A in Panther Lake. However, this does not mean that SMIC has a better process than Intel 18A or TSMC N3P. M0 is a local intra-cell routing layer. Its usefulness depends on the full interconnect stack: M1 and M2 pitch, track count, via and line resistance, design rules, mask count, overlay control, and routing flexibility. The 32.5 nm M0 is consistent with self-aligned quadruple patterning (SAQP), whose four-population line-width loading we read coarsely as alternating widths of 21.5 to 24 nm; M1 and M2, at 38 and 40 nm, are consistent with self-aligned double patterning (SADP), a single A/B split. On TSMC N6, M0, M2, and M3 sit at a relaxed ~40 nm and are consistent with SADP-class double patterning, with no need for quadruple patterning. That said, we measure M2 for example at ~43 nm, likely inflated by sparse routing. We do not assign any specific layer to EUV from our cross-sections; the distinction we can draw is double versus quadruple patterning, not lithography wavelength. Transistor-level density in the front-end-of-line (FEOL) sets an upper bound, but the design is ultimately limited by what the interconnect stack can route. The lowest metals are the most important for standard-cell density, but the semi-global and global layers determine how usable that density is at the block and chip level. Two axes are commonly used for chip cross-sections: the fin-cut and the gate-cut. The micrograph above is a fin-cut and shows metals 0 through 3. This axis lets us see the even-numbered metals, with M0 right above the fins. There are two kinds of M0 lines. The first are the power rails; these are wide wires for the VDD and VSS running horizontally at the top and bottom edges of each standard cell. The wide wires measure 55 nm across, more than double the other M0 lines. Their width minimizes resistance and reduces IR drop. The second kind are intra-cell wires, short segments within the cell that connect terminals to M1. These have alternating widths between 21.5 and 24 nm. The M0 pitch is 32.5 nm, a 19% reduction versus N+2 and N6. At this pitch, DUV patterning requires more aggressive multi-patterning, increasing mask count, overlay sensitivity, process complexity and cost. M0 is below what a single DUV-defined spacer (SADP) can resolve, so SMIC cascades a second spacer step (SAQP). The cross-section reflects the cost: the M0 trenches are visibly more re-entrant (narrower at the bottom than the top) than M1 or M2 on the same chip and carry a bright barrier-rich foot where the trench meets the etch-stop layer. That shape is partly the intended damascene profile, as a slightly narrow bottom helps void-free copper fill, but its magnitude at M0 is driven by the tight pitch and the higher trench aspect ratio. Intel 18A supports an M0 pitch of 32 nm, although Panther Lake has only shipped with a looser 36 nm pitch. This is due to Intel’s heavy usage of HP libraries. Among leading-edge nodes, 18A has the loosest M0 pitches due to PowerVia. With power routing moving to the backside, congestion is reduced, and the entire front-side metal stack can be used for signal routing. M2 is the first true inter-cell routing layer. It runs horizontally like M0 but spans across multiple cells to carry block-level signals. The M2 pitch sets the cell’s track height – the number of M2 tracks that fit between the VDD and VSS rails, defining what the library calls a 6-track or 7.5-track cell. This layer is the most important, limiting the routing of entire blocks. SMIC N+3 features a 5.7-track cell. The M2 pitch is 40 nm, a 5% decrease over N+2 and the same as N6. This shrink keeps the pitch at the edge of what is possible with double patterning. Future nodes will need to increase the number of masks for M2 as reducing the number of tracks is much harder due to the limitations in routing. The micrograph above is in the perpendicular direction, the gate-cut, and shows metals 0 through 4. This allows us to see and measure the odd-numbered vertical metal layers. The M1 pitch is 38 nm, 9.5% less than N+2 and 33% less than N6. The M1-to-gate ratio matters because it sets local routing flexibility. N+2 and N+3 use a 3:2 ratio, while N6 uses a 1:1 ratio, explaining the huge differences in M1 pitch. The more M1 lines there are compared with the gates, the more flexibility there is for power and signal crossing within the cell. Routing flexibility enables more complex and better cells. Clean fractional ratios are also preferred as a grid is periodic and improves layouts. The 3:2 ratio gives SMIC more local routing flexibility than a strict 1:1 grid, but it also complicates layout and patterning. This is a DTCO choice, with SMIC increasing process complexity to recover density and routability without EUV. This 3:2 ratio is not very popular in the leading-edge nodes. TSMC has only used it on N7+, the N5 family, and the short-lived N3(B). They have switched back to a 1:1 ratio for N3E. Intel has only used it on the 10 nm/Intel 7 family, with Intel 4, 3 and 18A all using a 1:1 ratio. Samsung is the only one still using a 3:2 ratio at the leading edge, using it in the SF4 and SF3 families. It remains to be seen if SMIC will remain at a 3:2 ratio or move to a 1:1 ratio with its future nodes. The industry is still actively exploring these local-routing ratios. At VLSI 2026, imec will be presenting work on even higher ratios, including a 2:1 scheme that can reduce area by up to 14%. We will be covering the conference in a future newsletter article. Subscribe We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline or to commission a custom teardown, contact sales@semianalysis.com. Interested in joining us on this ride and think you can be a difference maker? Check out our Careers page. The final local interconnect layer for N+3 is M3, with a pitch of 44 nm. The M3 pitch is the same as on N+2 and 10% larger than on N6. The semi-global layers carry the majority of block-level signal routing. They have a coarser pitch than the lower local layers. On leading-edge nodes, they are designed to sit at the limit of DUV single patterning. M4 through M11 pitches were found divided between 80–82 nm (M4–M6), 128 nm (M7–M10), and 148 nm (M11). Given limited sampling, it is possible these are divided further in dense routing areas. At the top are two giant metal layers, M12 and M13. These have kept the same pitches as N+2 at 1920 nm and 4600 nm respectively. While the lower layers’ pitches are generally fixed by the process and library, the upper layers vary much more in pitch and count, depending on the design. Even two smartphone SoCs on the same process can have wildly different metal stacks. The Helio G99 carries fewer routing layers, reaching coarse metal pitches of 850 nm by M9, while the larger and higher-performance Kirin 9030 keeps fine pitches until M11. SRAM At the leading edge, SRAM is much more difficult to scale than logic. TSMC’s latest nodes have seen little to no bitcell scaling, while logic still has more DTCO levers to pull. While looking for other logic libraries in the GPU compute units, we stumbled upon the SRAM. The most common type of SRAM has 6 transistors (6T), but this cell had 8 transistors (8T) instead. 8T SRAM adds two transistors to form a dedicated read port. Unlike a 6T cell, where reading disturbs the storage, the decoupled read port removes read-disturb, improving read stability and letting the cell be pushed harder for performance. At first glance, the cut looked like an unusual logic library, with each cell row having 3 fins of one polarity and 5 fins of another. The rows also alternated in orientation. Energy-dispersive X-ray spectroscopy (EDS) resolved our confusion. The cut had not landed on the GPU logic, but on the SRAM macro beside it. The unusual fin pattern was due to the SRAM library. We return to EDS in the process flow analysis behind the paywall. SRAM libraries are not like traditional logic libraries. Due to the unequal number of PMOS and NMOS transistors, they require specialized rules and layout libraries. They do not need the flexibility of logic libraries, so they are hyper-optimized for one purpose: dense, reliable memory. The SRAM cell we found is a 1:2:2-2:2 cell. This means there is 1 fin per pull-up (PU) PMOS transistor, and 2 fins per pull-down (PD) and pass-gate (PG) NMOS transistor. These 2 PU, 2 PD and 2 PG transistors would usually form a single 6T high-current cell (HCC). An 8T HCC adds a read-pull-down (RPD) and a read-pass-gate (RPG) NMOS transistor, each with two fins. We measured a cell height of 406 nm, which brings the bitcell size to 0.0463 µm². That is a theoretical peak density of 21.6 Mib/mm². We estimate that a 6T HCC would have a cell height of 292 nm and a size of 0.0337 µm². This is ~12% larger than a 6T HCC on Intel 3 and 4. We also estimate the 6T high-density cell (HDC) to have a cell height of 228 nm and a size of 0.0260 µm². This is coincidentally the same as the logic standard-cell height measured earlier. The estimate puts the cell near Samsung 7LPP/5LPP and slightly below TSMC N7/N6. That is a theoretical peak density of 38.5 Mib/mm². 6T HDC is arguably the most important cell as it is used for the largest caches in a chip, the L3 caches and system-level cache (SLC). Both the Kirin 9020 and 9030 have split the SLC into 4 banks to raise total SLC bandwidth. In the Kirin 9030, the SLC increased from 2 MiB to 3 MiB per bank. Correspondingly, the number of arrays within the bank also increased by 50%, from 16 to 24. Each array can store 128 KiB and forms an orderly pattern on the die shot. From the Kirin 9020 to the Kirin 9030, the area of a 128 KiB SLC array decreased from 0.0477 mm² to 0.0392 mm², an 18% shrink. The achieved density is 25.5 Mib/mm², 66% of the theoretical maximum. While the SLC was quite similar across both chips, the L3 has seen some major changes, particularly in terms of its layout. The total capacity also went up from 10 MiB to 12 MiB. Much like the SLC, the L3 is also split into 4 banks. In the Kirin 9020, an L3 bank consisted of 16× 128 KiB arrays and 16× 32 KiB arrays. However, an L3 bank in the Kirin 9030 instead consists of 48× 64 KiB arrays. In the Kirin 9020 L3, a 128 KiB array was 0.0513 mm² and a 32 KiB array was 0.0154 mm². The size of the 128 KiB array is different on the L3 and SLC as the assist circuitry for the two arrays differs depending on their purpose. In the Kirin 9030 L3, a 64 KiB array is 0.0210 mm². Although not a like-for-like comparison, normalized for capacity, it is 18% smaller than the 9020’s 128 KiB L3 array and 31% smaller than its 32 KiB L3 array. The achieved density is slightly lower than the SLC, at 23.8 Mib/mm², 62% of the theoretical maximum. Unlike the L3 and SLC, the prime cores’ private L2 cache uses a 2-bank design. As the prime cores’ L2 is latency-critical, it likely uses 6T HCC instead of 6T HDC. The 9020 has 16 arrays in each bank while the 9030 has 32. Each array has a capacity of 32 KiB. A 32 KiB array in the L2 shrank from 0.0171 mm² to 0.0142 mm², ~17% smaller. The density is 17.6 Mib/mm², ~59% of the theoretical maximum for 6T HCC. SRAM scaled well from N+2 to N+3, shrinking by ~19%, close to the theoretical logic shrink. The caveat is that N+2’s bitcells were unusually large, bigger than comparable 7 nm-class nodes, so part of the gain is catch-up rather than true scaling. With the insights from STEEL’s teardowns, we will be doing a deep dive into SRAM in a future newsletter article. Subscribe Everything above came out of a single STEEL teardown: die annotation, block-level area analysis and TEM cross-sections through logic and SRAM. We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline or to commission a custom teardown, contact sales@semianalysis.com. Future Roadmap The same cross-sections that pin down N+3 also show where SMIC can go next. Although N+3 is already close to the practical limits of DUV multi-patterning in several layers, SMIC still has a few scaling levers left. A theoretical N+4 would likely start with cell height. N+3 uses 5 M0 tracks between its power rails. Moving to 4 M0 tracks, as on SMIC N+2 and TSMC N6, could reduce cell height by roughly 15%. The routing grid is only one side of the shrink; the front end also must fit into the smaller cell. One possible FEOL lever is reducing the p-to-n isolation spacing from two diffusion grid units to one. Intel used this scaling booster on Intel 4, and TSMC did so on its N3 family. This path trades layout flexibility for density. Fewer M0 tracks reduce local routing resources, while tighter p-to-n spacing raises integration and design-rule difficulty. M2 is also constrained by the cell height shrink. For SMIC to maintain a ~5.7-track cell, M2 would need to move toward ~35 nm. That would move another layer into SAQP territory. SMIC could also reduce the CGP from 57 nm to 54 nm. Intel reached a similar CGP on Intel 10 nm/Intel 7 without EUV. The local interconnect is also tougher. If SMIC keeps the 3:2 M1-to-gate ratio, M1 would need to shrink to 36 nm and would likely require SAQP as well. If SMIC moves to a 1:1 ratio, M1 could relax to 54 nm, but it would give up routing flexibility. Under this theoretical path, we estimate that SMIC N+4 could reach a cell height of 198 nm and a CGP of 54 nm, implying a Bohr density of 137.8 MTr/mm², on par with TSMC N5 or Samsung SF4. However, the difficulty is cumulative. Each step is individually plausible, but together they make N+4 harder than the transition from N+2 to N+3. It will likely take longer, cost more, and carry less process margin. A theoretical N+5 would require a larger integration shift. One possible path is backside contacts (BSCon), moving power routing and source/drain contacts to the backside, which would reduce front-side routing pressure and enable another reduction in cell height. The front-side metal pitches could relax to reduce the process complexity. M0 would likely relax slightly to ~34 nm, and M2 and M4 pitches could relax further. CGP is unlikely to shrink much further. Even with EUV, 48 nm has been the practical limit for yield and process control. This approach would allow N+5’s cell height to fall to 170 nm and its CGP to 53 nm. This implies a Bohr density of 163.6 MTr/mm², on par with Intel 18A’s HP library. However, this would not make N+5 cost-competitive with the leading edge. It would reach a similar density through a much more expensive route. The integration difficulty rises sharply, with new process flows for backside alignment, wafer thinning, contact reveal, and backside metallization. Past this point, standard density and interconnect scaling become increasingly unattractive. That is where Huawei’s roadmap stops looking like a normal foundry roadmap and starts looking like a packaging roadmap. At ISCAS 2026, Huawei unveiled its tau (τ) scaling law, reframing process scaling in the time domain. τ is the time cost of data movement and processing: switching delays in transistors, RC signal propagation delays in circuits, compute, memory, and networking latency. Outside Huawei’s terminology, this is called system-technology co-optimization. This is Huawei’s answer to its lack of EUV lithography. Without EUV, planar density cannot keep pace with TSMC, Intel, or Samsung. If transistor density cannot shrink further, Huawei’s alternative is to shorten wires, reduce buffering, and stack logic vertically. “LogicFolding”, Huawei’s implementation of this new scaling idea, is, in practice, an aggressive 3D stacking approach. AMD V-Cache places SRAM above or below a CPU die. AMD’s MI350X places active interposer dies (AIDs) underneath accelerator and compute dies (XCDs), with the AIDs handling cache, IO interfaces, the network-on-chip (NoC) and embedded metal-insulator-metal (MIM) capacitors. With LogicFolding, parts of the same logic block are split across multiple active dies bonded face-to-face at ultra-fine pitch. This allows Huawei to shorten some critical paths and reduce buffer overhead, not merely add cache capacity or offload the IO and interconnect. Shortening wires is where the higher clocks come from. A large share of a modern core’s delay and energy budget goes into driving long interconnects and the repeater buffers along them. LogicFolding distributes a block’s critical-path gates across multiple stacked tiers bonded at a very fine pitch, so the bond interface behaves like an additional metal layer and the longest paths get shorter. That is how Huawei expects to recover frequency and efficiency it cannot get from the process alone. Huawei’s roadmap shows its intent. Prime core frequency is targeted to rise from 2.75 GHz in the Kirin 9030 to roughly 5 GHz by 2031, far beyond what planar scaling alone could deliver. Prime cores with 3.1 and 3.39 GHz clocks are being tested in its labs, although their power consumption is unknown. Beyond that, chips are in the design, simulation or pathfinding phase, meaning the frequencies are targets. However, the direction matters more: LogicFolding also helps with performance, not just density. The catch is that Huawei’s density claim is not directly comparable to foundry densities. A stacked design can report more transistors per package footprint by adding active layers, even if each patterned die remains well behind TSMC or Intel in front-end density. This is how Huawei can claim to reach foundry 14A-equivalent density by 2031. This is not a like-for-like foundry comparison, with Huawei using stacked logic and measuring density per package footprint. On a normalized Bohr-density basis, SMIC N+3 is ~114 MTr/mm², 38% less than Intel 18A’s HD library. Huawei’s 3D roadmap closes the gap by stacking active logic, reaching 215 MTr/mm² by 2030. In 2031, the roadmap density jumps to 295 MTr/mm², implying either a third active layer, partial EUV insertion or aggressive planar DUV scaling. Huawei’s methodology makes other foundries look much denser as well. Applying it to AMD’s MI450X with an N2 top die on an N3P base die yields a theoretical density of 460.2 MTr/mm² in 2026, compared with Huawei’s 295 MTr/mm² in 2031. This Kirin 9030 does not use LogicFolding, remaining in a conventional mobile SoC package. Instead, it forms the baseline for how far Huawei and SMIC can push planar scaling. Future teardowns of Kirin and Ascend chips will show both planar logic density and Huawei’s hybrid bonding solutions. Export controls changed China’s optimization problem rather than ending it. EUV restrictions raised the cost and complexity of leading-edge manufacturing without freezing it. SMIC reaches N6-class logic density through DUV immersion, SAQP and DTCO, while Huawei shifts more of the burden onto architecture, packaging and system-level integration. Future nodes will be tougher. N+3 still had room to tighten local metals and reduce cell height and CGP. Further scaling without EUV leaves fewer levers. More aggressive multi-patterning adds masks and overlay error. SMIC can keep pushing DUV, but each step will get more expensive and less forgiving. The design side is just as critical. Huawei had domestic EDA tools and flows before the Kirin 9030, with the Kirin 9000s, 9010 and 9020 making that clear. Huawei was able to ship multiple consumer SoCs on SMIC N+2 and N+3 while cut off from the Western EDA stack. US export controls restricted EDA tools for advanced chips in 2022 but did not target tools for more mature chips. In 2025, the US government briefly placed much broader restrictions on EDA software from Synopsys, Cadence, and others, before lifting them less than two months later as part of a trade deal tied to rare earths. Huawei has been unable to access those tools because it remains on a US trade blacklist. That forced Huawei, SMIC and Chinese academic institutions to build their own tools and flows. Researchers at Peking University recently announced a prototype EDA tool for Huawei’s LogicFolding architecture, which requires a new flow to handle the multilayer layout and floorplan. This is not the same as replacing the full Synopsys or Cadence stack, but it shows where domestic EDA is headed: toward tighter co-optimization between architecture, process and packaging. These advances are also diffusing into the Chinese ecosystem. SMIC is licensing its N+2 and N+3 processes to HLMC/Hua Hong at the government’s direction rather than by choice. If the same process learning feeds into Ascend accelerators for AI training and inference, the choke point shifts from one named fab to an ecosystem. Alibaba’s T-Head silicon arm and Cambricon, a Chinese AI chip designer that is expected to supply ByteDance, could also be major beneficiaries. Sanctions aimed at SMIC alone become less effective once the manufacturing knowledge has spread to other fabs and design houses. China is not closing the gap with Intel, Samsung and TSMC. The teardown shows the opposite in several places: no EUV, no backside power, higher process complexity, and visible trade-offs. But China is still advancing. If domestic chips become good enough for phones, inference, networking and security-sensitive workloads, they can matter strategically without matching TSMC at the leading edge. Behind the paywall, we show what else STEEL can do, with material and process flow analysis of SMIC N+3, and analysis of the Kirin 9030 package. We’re diving deep into the most advanced datacenter and AI hardware hitting the market. To learn more about what’s in the pipeline, access the full Kirin 9030 and SMIC N+3 analysis or to commission a custom teardown, contact sales@semianalysis.com.
Emma Raducanu loses Queen's final to Donna Vekic despite battling display by British No 1
Emma Raducanu was beaten by Donna Vekic in the womens singles final of Queens despite a battling display by the British No 1.
Dodgers star Yoshinobu Yamamoto loses perfect game on Mookie Betts error in 8th inning, then loses no-hitter in 9th
Sandy Koufax remains the only Dodger to throw a perfect game.
Browns front office loses a top executive to the Vikings
The Cleveland Browns are losing VP Andrew Healy to the Minnesota Vikings front office.
Fulshear's Xscape Theatre closes as EVO Entertainment plans 70,000-square-foot bowling, arcade and movie venue
Fulshear's Xscape Theatre on FM 1093 has closed permanently. New owner EVO Entertainment plans a 70,000-square-foot venue with bowling, arcade games and movies by summer 2027.
Lightning star Nikita Kucherov wins MVP trophy over Connor McDavid in closest vote in 24 years
Three different players got at least 50 first-place votes each.
Ubisoft Closes Belgrade Studio, 100 Employees Let Go
Ubisoft has closed its studio in Belgrade, Serbia. The shutdown will result in all of its employees being let go.
Hilo business closes storefront amid rising prices
A survey of Hawaiʻi businesses revealed that tariffs and trade policies have affected 86% of local businesses in the last year.
Pratt loses in LA mayor race, Trump stokes conspiracies
Pratt lost ground with every new batch of vote dumps
Iconic Bardstown Road business closes after nearly five decades
Electric Ladyland, an iconic Bardstown Road business, which has been in operation for nearly five decades in Louisville, has closed its doors.
Iconic Bardstown Road business closes after nearly five decades
Electric Ladyland, an iconic Bardstown Road business, which has been in operation for nearly five decades in Louisville, has closed its doors.
From Classroom to Racetrack: How Charlie Baur is Building a Motorsports Career in a Business Discipline
Charlie Baur can trace his racing story back to a Christmas morning in Michigan — the kind that starts with snow on the ground and ends with a throttle in a kid’s hands. “I remember being 3 years old and then for Christmas that year, my parents had gotten me a 120 cc snowmobile,” Baur said. Long before sprint cars, he was a Michigan kid on snowmobiles, dirt bikes, four-wheelers and jet skis — and a regular at car shows and races with his family. Photo courtesy of Charlie Baur That early exposure became a calling. At 12, he sat in a sprint car for the first time and knew he’d found the next chapter. He started in quarter midgets, moved to go-karts, and the day before he turned 16 — one day shy of the minimum age required by track insurance — he made his first sprint car start after convincing officials to let him race. Five years later, Baur is still pushing forward, now racing sprint cars on both pavement and dirt while pursuing an automotive business education at Northwood University — with a clear-eyed understanding that motorsports encompasses more than what happens under the lights on Saturday night. “It’s a business,” said Charlie, a 20-year-old who calls Fenton, Michigan, home when he’s not at a racetrack or at Northwood’s residential campus in Midland, Michigan. Photo courtesy of Charlie Baur Racing is the product — but preparation is the job To the average fan, race day begins around dinner time. For a team, race day can be a 12-hour shift — plus the work that came before it. And Baur’s racing operation is a family affair. “The team is my parents and myself,” he said, explaining they own the cars, maintain equipment and manage the business side — from sponsor relations to budgeting. On race weekends, the core group includes Baur and his parents (Eric and Stefanie), plus a couple of additional hands (Alex Photiou and Robert Flannigan) who help with the intensity and pace of trackside work. That structure has shaped how he thinks about efficiency and professionalism. “We’ve always kind of had the mentality of we’re never working on the car for maintenance at the track,” he said. “Our goal is to do as little work as possible to the car because it should be ready when that car rolls out of the trailer to go racing.” That mindset shows up in routines: bolt checks, tire procedures, safety equipment and a repeatable process designed to limit errors. “Once I fire the car up and put my helmet on, that’s when it’s go time,” he said. “You can plan as much as possible for how the race is going to play out, but until you take the green flag … you don’t know how it’s going to go.” Photo courtesy of Charlie Baur First-generation driver, student of the game In racing, “first-generation” matters — and Baur wears that label with pride. “A lot of the drivers are third or fourth generation,” he said. “My family has never been in motorsports. So for me, we’ve always looked at it as we’re new to it and we want to be as different as possible. From a marketing standpoint, it helps us stand out.” It also pushes him to seek out mentors — especially veteran drivers who have seen every scenario, solved every problem and won in multiple eras. Baur credits accomplished driver Jason Blonde, a 10-time champion, for becoming his driver coach and helping him develop as a competitor. “To me, learning from Jason helps me because he has had lots of success and knows what to do,” Baur said. “You have to humble yourself and say, ‘I don’t know what to do; can you help me out?’” That willingness to learn extends off the track, too. Baur says one of his biggest challenges is not a mechanical limitation, but credibility. “It’s my age,” he said. “Some people won’t take me seriously because I’m so young. Even though … I do know what we’re talking about and have a general idea.” His answer has been to out-prepare, out-communicate and treat the operation like a real enterprise — because it is. Photo courtesy of Charlie Baur Northwood’s role: Turning passion into a plan Baur is majoring in automotive aftermarket management and pursuing an accelerated path that pairs undergraduate study with graduate work. He’s also exploring how to align his studies with motorsports-focused coursework as Northwood advances related programs and pathways. He sees the value clearly: motorsports needs business-literate professionals who understand budgets, sponsorship, marketing, operations and long-term strategy — and who can translate between the language of the shop and the language of the boardroom. That’s why his Northwood experience isn’t limited to the classroom. He has been involved with the Northwood University International Auto Show and serves as motorsports chair — a leadership role that demands outreach, contracts, coordination and a steady professional demeanor. “There’s so much outside work that makes the auto show successful that people don’t realize,” he said. “And it is truly student led.” He’s also participated in industry experiences through Northwood at major events such as AAPEX, SEMA and PRI, along with other automotive networking events — opportunities he describes as a fast track to understanding how business really gets done. “You learn a lot in the classroom,” he said. “But if you learn it and you can’t apply it to the real world, it loses its value. Or if you learn it and can apply it, but you don’t know anyone, it is still of no value.” In his words, Northwood helps him win because it “gives me the tools to succeed” — including flexibility and faculty support as he balances travel and racing commitments with academics. Photo courtesy of Charlie Baur Racing as an ecosystem — and a talent gap waiting to be filled Baur is blunt about what he sees across the motorsports world: a deep bench of experienced professionals, and not enough young talent entering the pipeline. “I go to car shows, I go to these different events, and typically … the average age is between 45 to 55,” he said. “There’s not a lot of young people going into the motorsports world.” That reality reinforces Northwood’s communications focus: positioning motorsports not as a narrow lane, but as a wide ecosystem of business careers — sponsorship, logistics, data, media, leadership, hospitality and innovation — with real demand for prepared graduates. It also shapes how Baur talks to younger students who love racing but aren’t sure what their future looks like. “If you have a passion for motorsports, there’s gonna be a spot for you somewhere,” he said, emphasizing that driving is only one of many career paths — and that local tracks and hands-on experience are often the best first step. But if someone is interested in driving as a career path, it will take a team to make that happen. Baur’s team includes Stahl’s Motors and Music Experience, Rhino Racing, Randy Wise Buick GMC Chevrolet, Western Michigan Fleet Parts, Mother’s Polish, Schaeffer’s Oil, Hepfner Racing Products, Ameriguard Storage Centers, Justice Brothers, Race 1 Wraps, TBM Brakes, Behind Your Design, CharlieBaur.com and Northwood University. “There are so many people who invest not only in myself but in my race team and who believe in us to succeed,” he added. Pictured: Charlie Bauer’s sprint race car. Looking ahead: A headline worth chasing Ask Baur to imagine 2035, and he’ll give you two answers — one dream, one blueprint. “In a perfect world,” he said, “winning the Indy 500.” Realistically, he wants to build a life where business ownership and racing reinforce each other and where he can pay forward the help he has received by mentoring younger drivers, especially those trying to break into motorsports without a family legacy. “When I retire from full-time racing, I want to be able to provide the same opportunities that were presented to myself to the next generation of drivers,” Baur said. “I know how difficult it is to get your foot in the door.” For now, the focus is simple: keep building the operation, keep learning the business, and keep showing up prepared — because in sprint car racing, as in life, process matters. And once the helmet goes on? “It’s go time,” Baur said. Photo courtesy of Charlie Baur Learn more Students interested in motorsports careers — whether in competition, team operations, marketing, finance, sponsorship, data, or management — can explore opportunities through Northwood University’s Center for Automotive and Mobility Studies. For Charlie Baur’s latest motorsports adventures, follow Charlie_Baur and Charlie Baur Racing on social media. To learn more about Northwood’s Bachelor of Business Administration in Automotive and Mobility Studies and motorsports management concentration, visit www.northwood.edu/programs/automotive-mobility-studies/. Photo courtesy of Charlie Baur Pictured: Charlie Baur at Butler Motor Speedway on May 31, 2026. Share this page
Knapp’s, an iconic Tacoma diner, closes after nearly 90 years in business
Knapp’s Restaurant and Lounge has closed, according to a sign posted on its front door on Monday. The Proctor diner and bar opened in 1938.
We can create the future of science right now
Paul Litvak is the founder and Executive Director of the Robyn Dawes Institute and a Visiting Scholar at UC Berkeley. He has a PhD in Behavioral Decision Research from Carnegie Mellon where he studied emotions and the sunk cost bias. Over 15 years in industry, he solved a wide range of challenging technical problems. At Meta he created models optimizing the ad review process. At Google he ran experiments to measure social influence. At Airbnb, he was a product manager leading machine learning teams optimizing search ranking and price suggestions. He also co-founded and led product at Rhythmic Health, a venture-backed biosensing startup, creating an accurate low cost system using a color changing strip and a smart phone to measure salivary lactate. The bottleneck At this point, it is uncontroversial to say that science needs to stop using the PDF article as the unit of knowledge and currency. The unbearable slowness of scientific publishing, the profit motive and margins: I’m not saying anything new. The PDF also sucks because it’s hard to extract structured information, which makes it hard to do evidence synthesis. As a result, we do much less evidence synthesis than is needed. And evidence synthesis ultimately undergirds most policy and medical decision-making. I can see second-by-second real-time odds for any sporting or newsworthy event, but a school board can’t see the best evidence on whether their 8th graders should be taught algebra. As a society, we don’t treat this as an important problem. Again, not controversial. Not only is the problem well understood, but the solution has already been laid out. What we need is AI-assisted living evidence synthesis - (1) an open knowledge graph of atomic claims, (2) claims linked to evidence, (3) assessment and synthesis of each piece of evidence, and (4) continuous updating with new data. What few realize (yet) is that the technical capacity to build this vision for a significant portion of science already exists. Not only that, scientists and startup teams are already building many of these components. I know this because I’ve been surveying the space and talking to many of the builders. There are some missing pieces: for example, evaluations of how well some of the components work. But at this point, most of what’s missing is a fully end-to-end working integration of all of these parts. In the rest of this essay, I’m going to lay out all the parts of a working living evidence layer for science and who is working on them, and propose concrete next steps for building this system. What’s now possible The diagram above outlines the components of a living evidence synthesis platform, including some of the teams working on each component1. Scientific PDFs are processed into claims with associated evidence. The evidence is subjected to a forensic audit, methodological evaluation and robustness and reproducibility checks. Finally it’s given a weight in a continuously updating synthesis. What follows is a description and status of each component and a few of the teams working on them. Document understanding The first thing you need to be able to do is turn an article into structured data. Mostly that means parsing PDFs. There are often multicolumn layouts that confuse non-specialized PDF to text processing libraries. For scientific papers, there is the added complexity of parsing formulas and tables and figures. This is a really hot area - there are startups offering APIs, and it seems like a new open source package gets posted to Github every few weeks. What follows isn’t exhaustive. A package called GROBID was the state of the art for a while, they didn’t update their package for nearly two years until very recently. In the meantime reducto.ai released an AI powered PDF extraction API, PaddleOCR became popular, IBM released a model called Docling, and both Mistral and Gemini created models and libraries. I also know of at least one other well-funded psychology research group working on a paper parser. By contrast, there are few open evals in this space, with no extensive evals for complex table comprehension in particular. Nonetheless, I’m confident this will be a solved problem soon, given the combination of LLM advances and developer interest. Hypothesis level extraction There has been increasing interest in comprehending the extracted text of papers and linking information to evidence for each hypothesis. A lot of work has already been done. Trialstreamer (Marshall et al. 2020) and RobotReviewer LIVE (Marshall et al. 2023) demonstrated automated extraction of trial population, intervention, and outcome at scale on clinical RCTs. PaperQA2 (Skarlinski et al. 2024) and Ai2 ScholarQA (2024) extended this to retrieval-augmented question answering with citation grounding. Elicit, Consensus, and SciSpace operationalized claim-level extraction for end users. OpenEval (Booeshaghi et al. 2026) is the most recent and most ambitious: 1.96 million atomic claims extracted from 16,087 eLife manuscripts using Claude Sonnet 4.5, grouped into ~299,000 results, with LLM evaluations showing 81% agreement with human peer review on a 2,487-paper subset. None of these solutions link claims to test-statistics, as you would need to evaluate randomized controlled trials. This is why I built the evidence.guide API - to extract hypotheses and associated test statistics from behavioral science papers. The best public eval of this kind of extraction I’m aware of comes from the recent SCORE project - they had humans code thousands of psychology papers to extract their claims by hand. It would be extremely helpful to the world if all scientific PDFs were available as structured open data. I’ve been working to make this happen, both directly at Berkeley and through coordination with large entities I can’t yet speak of; as hard as it is to do, I think it’s possible2. Forensic audit A lot of work has been done on forensic audit, but some gaps remain. Of course, for biology papers that rely on images for evidence, there are a variety of tools (notably Proofig and ImageTwin) to spot anomalies. These are still well short of what sleuths like Elizabeth Bik can do on her own, but these tools are constant companions among fraud analysts. There’s someone working on auditing Excel files for anomalies, and a number of teams are automating numerical checks like GRIM and SPRITE, including Scrutiny project, the INSPECT-SR team as well as statcheck. The regcheck team is building a way to use AI to compare preregistrations to analyses in papers, to ensure there aren’t significant deviations. Nonetheless, there are many other kinds of anomalies to screen for, both public and less publicly known. And there are no formal evals for anomaly detection that I’m aware of. Still, there’s a lot to draw from in this space and I’m pretty certain we will be able to scan papers for most kinds of obvious anomalies in the near future. Methodological review This area has been white hot, though I fear for many of the startups in this space, because this capability may become commoditized. There are at least six different AI peer review companies, including Refine.ink, Reviewer3, ReviewerZero.ai, Q.E.D. Science, Paper Wizard, and Isitcredible. Coarse (a pun on refine) was also recently created as an open source alternative. These systems provide qualitative feedback on the content of papers, spotting methodological weaknesses and mathematical errors. They seem to work pretty well, and many academics report bitterly that they exceed the average quality of typical peer reviewers. But there are few evals here either. What evals exist so far involve using LLM-as-judge (circularity problems abound) or comparing against human reviews of questionable quality. What you’d ideally want is an eval that measures capturing known errors in papers3. Reproducibility and robustness Another active area has been using AI agents to automate computational reproducibility4 and robustness5 checks in papers that report numerical results. For more recent papers where data and code are available, AI agents can see whether they can re-run the analyses and produce the numbers reported in the published paper. In addition to a handful of individual academics who have been experimenting using Claude Code for this, the Institute for Replication is a leading group working on building an end to end system. The evals related to this problem are the most mature, with CORE-Bench (Siegel, Kapoor, Narayanan 2024) and PaperBench (OpenAI 2025) available to benchmark agents on this task. There is also work on getting AI agents to test alternative ways of analyzing the data to ensure the results are robust to small analytic design choices. Synthesis This is the most underdeveloped area where significant investment is required. Although some automated evidence synthesis systems exist — for example, otto-sr is building an AI agent to write systematic reviews — none of these incorporate the full range of paper level signals to weight evidence appropriately. Nor is there anything like an eval or a gold standard for a good systematic review. Arguably Cochrane reviews are the closest we have to gold standard human systematic reviews, though I’ve heard academics in the know complain about their uneven quality. A key question for a synthesis platform is how to weight anomalies and methodological issues in assessing the quality of a piece of evidence. This is an unsolved problem and one I’m very keen to work on. Continuous updating There are many pieces of basic infrastructure available for monitoring for new research and initiating updates. OpenAlex is the current open citation graph. Retraction Watch integrated into Crossref in October 2023. Scite tracks how citations support, contrast, or mention prior claims. The Living Evidence Network demonstrated continuous-update workflows in clinical guidelines. Engineering this is a relatively straightforward task. When you look over this technical architecture and all the progress being made, it’s hard not to be optimistic that a living guide to scientific evidence will be built. The stakeholders are ready The social infrastructure for this is starting to coalesce — it’s not just a pie in the sky academic exercise to imagine this coming into existence. Institutions like the Center for Open Science, the Institute for Replication, the INSPECT-SR, the Living Evidence Network and more are all working on scaling work to improve research quality. Funders are also aligned. The Sloan Foundation has funded living evidence work through COS. Coefficient Giving supports the Institute for Replication and COS. The Astera Institute and the Institute for Progress have shown interest in this space. NIH has established an Office for Replication and Reproducibility. Although there are (very unfortunately) serious headwinds in science funding generally, there is an active group of funders interested in metascience. A brief word about what I’ve been doing at RDI. First, as a Visiting Scholar at Berkeley I’ve been actively figuring out how a non-profit and a public university can conduct and make public the results of large-scale academic article data mining. With some of the money I raised from donors, I commissioned a legal analysis of recent case law and publisher text data mining (TDM) agreements in order to understand whether a massive open data mining of academic articles is possible (with caveats, it is). I’ve also been working to bring together stakeholders in this space, and identify gaps. I’ve also been doing some software development in this space, with more to come. A pilot proposal The assumption undergirding all of this is that an AI, given all this information, would make the right judgment about a scientific claim with lots of conflicting evidence, weighing all the factors appropriately. That’s the hypothesis we need to test. Randomized control trial research is the best place to focus on first. RCTs are used to make many of the important decisions in society - from medical trials to public policy changes. And they use a relatively uniform set of inferential statistics with lots of known and available diagnostics. Behavioral science experiments, within the broader realm of RCTs, should be first used as a testbed whose results can be generalized. Because behavioral science is at the vanguard of open science practices, replications abound (there are thousands of them) to serve as ground truth training data. Key Hypothesis Therefore the pilot would test, in behavioral RCTs that have been replicated, whether the quality of evidence for a claim can be used to accurately predict whether that claim will replicate. Secondary Hypotheses Compared to claims that replicated, non-replicated claims demonstrate a greater share of forensic anomalies in their source literatures. Hypothesis level claim and statistic extraction is accurate enough to scale living evidence without onerous human review costs. Replication prediction is more accurate than prediction markets6 or journal prestige. If all the different quality signals we gather do accurately predict which studies will replicate, then we can use that model to score evidence to power the living evidence layer. Why this is informative regardless of outcome If the pilot succeeds, the architecture extends to medical RCTs (where Living Evidence already operates and integration is mostly about claim representation), then to slices of basic biology with stable replication structure. If it fails, the field learns which quality signals are load-bearing and which ones metascience has oversold. Either result is a contribution to knowing what the literature supports. Implications for funders Because this burgeoning ecosystem of builders already exists, a well-informed philanthropic or government funder could play a crucial catalyzing role in bringing this future about. They could play at least three roles: creating open structured datasets, publishing open benchmarks and incentivizing the solving of key technical challenges. First, open archives of papers that the government maintains, like PubMed, could be turned into structured data amenable to large scale metascience and claim aggregation. I know the US government already has an interest in doing this, though some key open questions remain unanswered. How do you determine the best models and systems for accurately extracting information from papers? How can you establish a robust way to allow researchers to flag errors and correct them? And finally, how do you create a legal regime in working with publishers to maximize the scope of available papers? For the latter, a university or a private philanthropy may be better positioned to make structured data publicly available under journal subscription terms or fair use. Philanthropists or government funders could also coordinate to create or commission benchmarks that evaluate whether important problems have been solved. For example, an open benchmark for claim extraction from a range of different scientific article types would be extremely helpful. Ensuring the underlying data are accurate is vital for creating these evaluations. I’ve discussed opportunistically using various human-created datasets for this purpose, but a consistent problem is that errors in human data make it difficult for them to serve as a gold standard. Finally, with structured data and benchmarks available, the government or private philanthropy could use them to incentivize groups to develop machine learning systems that meet these benchmarks. Prizes are one potentially valuable tool for this. For example, you could establish a prize for a system that accurately updates a living meta-analysis for a small set of claims. Prizes are particularly useful as signals of problem importance, and can help create vibrant ecosystems of public and private research—see, for example the role the government played in kickstarting the current work on nuclear fusion. Conclusion The drawbacks of the current scientific publishing system are known. Scientists agree, metascientists agree, philanthropists agree: the published PDF plus citation graph isn’t the right substrate for maintaining a representation of the evidence base in science. The pieces needed to build the alternative either already exist or are rapidly taking shape. The community is forming around exactly this problem, with concrete partnerships and shared infrastructure. A pilot should start on behavioral science RCTs because that’s the slice of empirical science most amenable to legibility, where replication ground truth is richest, and where the failure modes are best documented. What’s been missing is the galvanizing mission to assemble these pieces into something that works. That’s what I’m proposing to build. *** 1 I have a broader field map that I’ll release publicly soon. This is me, building in public! 2 If this is something that you are excited about, please reach out and talk to me. 3 More on this very soon too! 4 This tests whether, given the code and the data, you can get the same statistics as reported in the published paper. 5 This tests whether the results are the same as a paper’s given alternative analytical decisions in conducting the analysis (like outlier omission). Closely related is the idea of a “multiverse” where you come up with many different ways of answering the same underlying research question with the same data, and test whether the results hold in all those alternative methods. There’s been work on the latter as well. 6 Some of the replications, e.g. those from the SCORE project, had paired the experiments with forecasts from prediction markets. So we get to look at this for free.
Kansas City entertainment restaurant, bar closes. ‘Difficult decision’
“This was an incredibly difficult decision, but one that I ultimately felt was necessary for myself and my family,” the owner wrote.
Kansas City entertainment restaurant, bar closes. ‘Difficult decision’
“This was an incredibly difficult decision, but one that I ultimately felt was necessary for myself and my family,” the owner wrote.
Planetarium show explores the science of the sun and moon
A live presentation at Samford University’s Christenberry Planetarium will examine how Earth’s closest celestial neighbors continue to shape life on our planet.
Spencer Pratt, former reality TV star, loses ground in LA mayor's race
Former reality TV star Spencer Pratt has lost significant ground in his bid to become mayor of Los Angeles.
With qualifying days away, 5 competitive House seats remain unchallenged
Palm Beach County’s House District 89 is the closest of the bunch, with 2024 presidential results showing just a 1-point Democratic advantage.
L.A. Mayor's Race: Nithya Raman Closes in on Spencer Pratt
Nithya Raman has overtaken Spencer Pratt in the race for L.A.’s next mayor.
New dummy units give our closest look yet at the iPhone Fold
We don’t know yet whether Apple will call it the iPhone Fold, iPhone Ultra, or something else — but everyone expects...
Dodgers on Deck: Sunday, June 7 vs. Angels
The Dodgers finish off their shortest homestand of the year with their third and final game of the weekend against the Angels at Dodger Stadium. Emmet Sheehan closes out his two-start week by starting in the series finale, after lasting at least six innings in three of his last four starts. José Soriano starts for […]
Steak n Shake says it ‘remains undefeated in political endorsements’ after Trump-backed candidate loses in Iowa
Fast food chain Steak n’ Shake is touting its political prowess after the Republican candidate it endorsed for Iowa governor defeated a challenger backed by President Trump. The restaurant posted o…
Cubs CF Pete Crow-Armstrong loses ball in lights to gift A's inside-the-park HR, then finds redemption with walk-off
The Cubs standout posted a rare defensive gaffe, a homer and a walk-off single.
California loses Fortune 500 crown to Texas as billionaire tax threat looms
Governor Greg Abbott calls Texas the "undisputed headquarters of headquarters" as the state surpasses California with 57 Fortune 500 companies.
Cam Heyward is still going strong as he closes in on the Steelers' record for longevity
If Cam Heyward keeps getting new contracts or years added to his deal, he will creep closer to a Steelers record not many players could ever sniff. Heyward’s...
Phillies ace Cristopher Sánchez loses scoreless innings streak at 50 2/3 innings, record among MLB lefties
Austin health agencies move to protect sexual health services as Safe Alliance closes
With the Safe Alliance set to shut down its sexual assault victim services next week, local public health leaders are moving to protect other critical sexual he
Pirates' Gregory Soto: Closes out eighth save
Pirates' Gregory Soto: Closes out eighth save
WhiteHawk Capital Partners Closes $390 Million Refinancing For The Star Entertainment Group
WhiteHawk Capital Partners, a private credit investment manager focused on asset-based financing solutions, announced the closing of a US$390 million senior secured financing for The Star Entertainment Group. The transaction provides additional liquidity to The Star while refinancing certain existing debt obligations.