SCROLL NEWS / DISCOVERY
Search headlines
Find the stories shaping the conversation.
Results for " gpus " 121 found 🔀 AI-powered Shuffle
AMD FSR 4.1.2 Now Runs On Intel Arc XMX Cores, Delivering Better Image Quality Than XeSS Upscaling With Similar Performance
Modding community brings AMD FSR 4.1.2 to Intel Arc GPUs, showing superior image quality and efficiency with upscaling times as low as 0.59 ms.
AMD Announces FSR 4.1 for Xbox Series X/S, Z2 Extreme and Z2A Handhelds, RX 6000 GPUs and More
Plus: FSR 4 will soon support Vulkan games.
AMD is bringing FSR 4.1 to Xbox consoles and handhelds
AMD’s latest upscaling tech is also coming to laptops, 6000-series GPUs, and handhelds.
AMD Ryzen AI Halo Review - Local-AI Powerhouse with 128 GB of Memory
The AMD Ryzen AI Halo is great at running local AI workloads, featuring a Ryzen AI Max+ 395 APU with 128 GB of unified memory in a compact form factor. It excels at handling large models, enabling tasks that most consumer GPUs struggle with. With AMD AI Playbooks, getting started is simple, allowing you to explore the frontier of AI at home.
New Tools Let Users Edit and Load Modified VBIOS on GeForce RTX GPUs Without Flashing
A developer who goes by Panchovix has released three tools that let NVIDIA GeForce RTX 50, RTX 40, and RTX 30-series owners run a modified VBIOS without writing anything to the card's flash chip. The first is a VBIOS editor for Windows and Linux that can change power limits, voltage ceilings,...
National Compute Grid Launches to Pool Idle AI Capacity Across Providers
For all the talk about a shortage of AI computing power, and the need for more powerful chips and bigger data centers, there’s another side to the story. A significant amount of existing capacity isn’t being used. While tech companies race to build massive new data centers, expensive GPUs in other facilities can sit idle […]
Report: RTX 5090 GPUs discontinued by Nvidia, as GB202 is reserved for AI workstation cards
Not that I could afford one anyway.
NVIDIA Reportedly Ending the RTX 5090, Reserves GB202 Silicon for Workstation GPUs
NVIDIA has reportedly stopped allocating GB202 silicon to the GeForce RTX 5090, according to leaker MEGAsizeGPU (@Zed__Wang) on X. "No more GB202 for 5090, confirmed," the leaker posted, adding in a follow-up that "all future GB202 supplies are RTX Pro exclusive." GB202 is the largest "Blackwell"...
NVIDIA CEO Jensen Huang receives National Medal of Science for turning gaming GPUs into AI powerhouses
NVIDIA CEO Jensen Huang received the National Medal of Science for advancing GPUs into general-purpose computing engines that power modern AI.
NVIDIA reportedly discontinuing RTX 5090, GB202 GPUs to be reserved for RTX PRO series
NVIDIA is rumored to discontinue its RTX 5090 lineup and introduce an RTX 5080 with 24GB. The claim comes from an unverified factory message.
NVIDIA-Backed Upscale Wants to Make Rival AI Chips Work Together
NVIDIA GPUs dominate today’s AI data centers, but they’re no longer the only processors companies want to use. Cloud providers are developing their own AI chips, while specialized accelerators are entering the market. Getting these different processors to work together efficiently is creating a new challenge for data center operators. That’s the problem Upscale hopes […]
Gears of War: E-Day GPU Benchmark: 44 GPUs Tested
We tested 44 GPUs in Gears of War: E-Day at 1080p, 1440p, and 4K to see what it takes to run the game well, how 8GB cards...
The Amazing Work By Valve's Timur Kristóf On Improving Old AMD GPUs On Linux
Over the past year Timur Kristóf of Valve's Linux graphics driver team has made multiple very nice improvements to the AMDGPU kernel driver for enhancing support for old (GCN 1.0/1.1 era from a decade ago) graphics cards so that they can better handle Linux gaming and other tasks
Gears of War: E-Day Performance Benchmark Review
Gears of War: E-Day is a prequel set 14 years before the original game. Built on Unreal Engine 5 with RTX Mega Geometry, DLSS 4.5, FSR 4, XeSS, and ray tracing, we examine graphics quality, VRAM usage, upscaling options, and how it performs across a wide range of modern GPUs.
Borderless Gaming’s DLSS 5 Implementation Runs Without Touching NVIDIA’s Software, Quantized to Work Well Even on Unsupported GPUs
Borderless Gaming has been updated with a custom implementation of NVIDIA DLSS 5 Neural Rendering that runs on a much wider range of GPUs.
Borderless Gaming adds DLSS 5 Neural Rendering without modifying game files
Borderless Gaming can now apply DLSS 5 Neural Rendering without replacing game DLLs or modifying game files. The quantized BGFX effect can run on GPUs including the RTX 3090.
The Witcher 3 Remastered Crushes GPUs With Path Tracing, but Free Mod Claws Back Up to 30% Performance
Following its surprise announcement during this year's gamescom Opening Night Live, The Witcher 3 Remastered is now available on PC, PlayStation 5, Xbox Series X|S, and Nintendo Switch 2. While the number of improvements to gameplay is staggering and well thought out, what most were looking forward to was the visual uplift, which includes very demanding path tracing support on PC. Thankfully, a mod has already come to the rescue, as it did after the launch of the Next-Gen update, optimizing both ray tracing and path tracing for better performance with minimal loss of quality. "Faster RT/PT in The Witcher […]
Not the GPUs. This Overlooked Piece of Nvidia's Business Could Matter Most for the Stock's Next Leg Up.
Nvidia sees this area as a $200 billion opportunity.
Not the GPUs. This Overlooked Piece of Nvidia's Business Could Matter Most for the Stock's Next Leg Up.
Nvidia sees this area as a $200 billion opportunity.
Modder Squeezes 74% More Performance Out of DLSS 5 on AMD Radeon in One Day
Modders working on bringing NVIDIA's DLSS 5 Neural Rendering to AMD Radeon GPUs just had their busiest day yet. The developer behind the DLSS-NR-on-AMD tool shipped two new versions within 24 hours, Alpha 0.3.3 and Alpha 0.4.0, together adding roughly 74% performance on Radeon hardware, according...
Intel Panther Lake Teardown, 18A, BSPD, GAAFET, SemiAnalysis STEEL
Panther Lake debuts the first commercial implementation of backside power delivery (BSPDN), introduces Intel’s first iteration of gate-all-around (GAA) transistors, and showcases their advanced packaging capabilities with its Foveros-S assembly. With Panther Lake, Intel’s manufacturing arc has shifted from nebulous roadmaps to shipped silicon, a significant milestone on their long road back to competitive semiconductor manufacturing. To evaluate the extent of Intel’s comeback, we tore down Panther Lake. The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page. Our teardown traces 18A from its four-sheet RibbonFETs (Intel’s marketing name for GAAFETs) and gate stacks through contacts, frontside and backside wiring, and the bonded carrier. We explain how these material and integration choices improve gate control and reduce resistance, while adding capacitance, thermal resistance, and process complexity. Our measurements put Panther Lake’s 18A compute logic and TSMC N3E GPU logic at similar logic density. However, 18A does not lead TSMC N3P, N2 or Samsung SF2 in peak density. Panther Lake’s CPU cores are incremental updates, and the high-end GPU still uses TSMC N3E. Panther Lake assembles one compute tile, one GPU tile, and one I/O tile atop a passive base tile using Intel’s Foveros-S advanced packaging. Both compute tile variants use Intel 18A. The Xe3 GPU options are a 4-core GT1 tile on Intel 3 and a larger 12-core GT2 tile on TSMC N3E. Both I/O tile variants use TSMC N6. [1], [2] Our analysis centers on the PTL-U compute tile, both the 4-core and 12-core GPU tiles, as well as the 12-lane I/O tile. In conventional chips, power and signal are routed through the same frontside metal stack towards the device frontend. Power rails consume scarce routing resources near the transistors, while tall via stacks carry VDD and VSS from the coarse upper wires to local rails. Backside power delivery (BSPD) moves the main power network behind the transistor layer, to the backside, separating it from frontside signal routing. We covered BSPD and its impacts in 2024. [3], [4], [5] Intel’s BSPD implementation, branded as “PowerVia”, routes power through dedicated backside metals to nano-TSVs, which connect those rails to local source/drain (S/D) contacts. Implementing that separation requires Intel to build the interconnect stacks from both sides of the wafer. The frontside comprises the M0-M14 signal stack, while the backside comprises the BM0-BM5 power stack. M0 and BM0 are closest to the transistors. The nano-TSVs connect the two sides, but Intel patterns and etches each via from the front after forming the contacts. A narrow via runs from the side of the contact deep into the silicon substrate. Intel then completes the frontside signal metal stack, bonds the wafer to a carrier, flips it and removes the original substrate until the buried via tips are exposed. The backside metal stack is then deposited directly on the revealed vias. The nano-TSV and backside-via profiles taper in opposite directions because Intel forms them from opposite sides of the wafer. The transistor structures form the FEOL. Local contacts and nano-TSVs connect them to the wiring. M0 begins the frontside interconnect stack. The silicon carrier remains attached above the frontside interconnects. It supports the device wafer during substrate removal and backside processing and remains part of the finished chip’s thermal path. PowerVia removes the main power distribution from the congested frontside metals, routing supply through shorter and wider backside wires. Its lateral landing still occupies area in the standard cell, so it recovers less cell area than a direct backside contact. [3] Nano-TSVs beside the logic devices carry VDD or VSS from the backside power network, while signal connections continue upward through the frontside metals. Backside Interconnects Samsung SF2 data is included for comparison to Panther Lake’s within this article. SF2 is the incumbent GAA foundry node but lacks BSPD, serving as a useful reference to evaluate 18A. A full teardown of Samsung’s S26 products, processed on SF2, will be shared soon.Nanosheet-cut EDS comparison. The PowerVia supply path runs from the backside Cu rails through Mo-lined W nano-TSVs to the local transistor contacts. In this cross section, the tapered connection spans roughly 150 nm from the contact level to BM0. The Ta liner confines Cu and promotes adhesion to the surrounding stack; the AlOₓ etch stop controls the next dielectric etch above the rail. Dielectric beneath the ribbons electrically separates the devices from the backside wiring and removes the conducting silicon body below the channel. [6] AlOₓ serves as an etchstop (ES), enabling endpointing and protecting the underlying layers. Low-volatility aluminum fluoride reaction products resist the fluorinated plasma, allowing a thin AlOₓ film to protect the metal while the surrounding low-k dielectric is removed. [6], [7]. While the BM0 and layers above the M1 lines show double AlOx layers, Our SMIC N+3 teardown showed single AlOₓ layers. SMIC uses a simpler local AlOₓ substack, while the remaining cap and etch sequence provide the required landing protection. So why double layers? The closely spaced AlOₓ doublets provide two protected endpoints in the etch sequence. Intel documents an AlOₓ/SiN/AlOₓ stack that explains the benefit. The main dielectric plasma etch stops on the first AlOₓ film; a selective wet clear opens that film; a second plasma etch removes the intermediate SiN and stops on the second AlOₓ film. The final wet clear exposes the metal landing surface. SiN is the intermediate dielectric in Intel’s published example. [8] The second stop protects the metal through a cap breakthrough. Wide openings can etch faster than narrow ones, and etch depth varies across the wafer. Metal under an early-clearing opening would otherwise be exposed while other openings still need more etching. Staged protection widens the process window and reduces metal erosion, corrosion and void formation. [8] TSMC documents AlN/AlOₓ/SiOC/AlOₓ above Cu, with AlN blocking Cu diffusion, and a simpler AlN/SiOC/AlOₓ variant that omits one AlOx film. [9] Levels with different opening sizes, aspect ratios, pattern densities and cap materials need different etch margins. A double AlOx stop is useful where another protected endpoint justifies the added processing. The extra film adds formation, selective opening and cleaning steps, plus another set of interfaces to control adhesion, moisture, and stress. These blanket films are opened through the existing via pattern, so each film does not require another lithography mask. AlOₓ adds parasitic capacitance when it replaces lower-k dielectric; two thin AlOₓ films can nevertheless contain less AlOₓ than one thick film. Total thickness, placement, and theintermediate dielectric determine the electrical cost. Deposition chemistry also changes AlOx permittivity and residual hydroxyl content, which can oxidize the underlying metal. [7], [8], [10], [11] The backside stack separates into relatively fine BM0-BM2 wiring near the devices and coarser BM3-BM5 power distribution. The largest pitch increase occurs between BM2 and BM3. BM0’s pitch closely matches the logic-row height, fitting local power delivery to the cell rows. Higher levels aggregate current through larger conductors: routing density becomes less important than low resistance and current capacity as the network approaches the package. This hierarchy provides wide power wiring for the power delivery network without consuming scarce frontside signal-routing resources. [3] The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page. Frontside Interconnects Intel 18A combines Mo-lined W contacts and nano-TSVs with a separate backside Cu power network. Samsung SF2 keeps power on the frontside, using Ti-based contact interfaces and Ta-based barriers and Co liners around Cu wiring. In 18A standard-cell rows, backside power rails supply the devices through nano-TSVs within the cells, freeing frontside routing resources. Samsung’s M0 accommodates both power and signal connections. From the device toward M0, the connection runs through a Ti-based S/D interface, W contact fill, a Mo-lined W via, and the Cu M0 wire. Mo supplies a conductive nucleation and adhesion layer for W, replacing the resistive TiN liner used in conventional W integration. This increases the effective conduction volume within the feature while retaining W fill and its established polishing, cleaning and etching processes. Intel’s Mo/W patent describes this integration tradeoff. The nano-TSV uses the same Mo-lined W construction in the backside supply path. [12] The move from TiN to Mo is an incremental change. While a full Co or Mo fill can also reduce the volume lost to liners in very small features, it requires new integration schemes that increase complexity and risk. Cu remains attractive for wider wires due to its low resistance. As wires and vias shrink, the diffusion barrier consumes an increasing fraction of their cross-section. [12], [12], [14] Intel uses Co/Ru liners at M0-M1, Co at M2-M4, and Nb at M5-M9. The lower-level liners help Cu adhere and reduce void formation during trench fills. Applied Materials’ Endura has new thermal control that facilitate wetting process, so the thin film continuity is good enough that good capillary pressure will drive Cu atoms to the via bottom without voiding. Intel’s choice to use Nb is particularly interesting. Intel’s Nb patent describes a conductive diffusion barrier intended to reduce the barrier’s contribution to resistance relative to conventional Ta-based barriers, particularly at via bottoms where all current crosses the barrier. The patent pairs Nb in coarser levels with the option of lower-cost PVD processing. [15], [16] The upper metal layers support thicker barriers formed through physical vapor deposition (PVD) despite its worse coverage and uniformity. Meanwhile, the lower metal layers require thinner barriers deposited through conformal atomic layer deposition (ALD). Co/Ru adds another material interface and requires controlled deposition and Cu fill. Changing liners and barriers by metal layer allows Intel to optimize interconnect resistance, process complexity, and reliability. [15, 16] RibbonFET, Intel’s name for its gate-all-around FETs (GAAFETs), replaces the FinFET’s vertical fins with four stacked horizontal silicon nanosheets, allowing the gate to surround the channel on every side. The path to GAAFET begins with the planar transistor. A planar MOSFET places the gate above the channel between its source and drain. Pairing an NMOS with a PMOS transistor creates a CMOS inverter, in which the NMOS pulls the output low for a high input, and the PMOS pulls it high for a low input. The gate must retain electrostatic control of the channel to ensure clean switching. As gate lengths shrank, the drain began to compete with the gate for that control, increasing off-state leakage. Electrostatic control was restored through an architectural evolution that raised the channel into a vertical fin and wrapping the gate around three sides. Called “FinFET”, this new architecture packed more effective channel width into a smaller footprint. Further scaling made it harder to maintain both drive current and leakage within smaller cells, and reintroduced the same problems planar MOSFETs faced. Nanosheet GAAFETs close the fourth side by replacing the vertical fin with a stack of horizontal nanosheets, each surrounded by the gate. The tighter electrostatic control suppresses leakage at shorter gate lengths while stacking adds effective channel width within the cell footprint. In a FinFET process, channel width changes in discrete steps as designers must add or remove whole fins. Nanosheet width can instead be adjusted continuously within the process’s design rules. Wider sheets increase drive current, while narrower sheets reduce capacitance at the cost of drive current. Intel 18A uses stacks of four nanosheets each and varies their widths across logic and SRAM. At the process level, adding more sheets to each stack increases effective channel width and drive current, but complicates fabrication. RibbonFET vs MBCFET Samsung began GAAFET production in 2022 with SF3E, following with SF3 and now SF2. Its ‘MBCFET’ provides a useful structural comparison with Intel’s first RibbonFET implementation. [17] STEEL is digging deeper into SF2, used in the Exynos 2600, and TSMC’s GAAFET N2, used in Apple’s A20 Pro, in upcoming newsletter articles. We’re throwing some teasers on X. Let’s compare Samsung SF2’s MBCFET with Intel 18A’s RibbonFET. Subscribe Even to the untrained eye, Intel’s extra nanosheet is obvious. Intel stacks four ribbons to Samsung’s three. Samsung’s sheets are much wider in these fields, so both sheet count and width matter to the available channel perimeter. Sheet width also changes which silicon surfaces carry current. On conventional (001) silicon, wide nanosheets emphasize the broad top and bottom surfaces, favoring electron transport; the larger sidewall contribution in a narrow sheet favors hole transport. Thinner sheets improve gate control but increase confinement and scattering. This makes width and thickness part of the NMOS/PMOS balance, alongside strain and threshold voltage. [18], [19] GAAFET designs like 18A use different work-function-metal (WFM) stacks for NMOS and PMOS. Around each ribbon, a thin SiOx interfacial layer separates the silicon channel from the HfOx high-k dielectric, with La providing dipole tuning and the WFM wrapping the dielectric. NMOS uses a TiAl-based stack, while PMOS uses TiN WFM. W fills the remaining gate trench, providing a lower-resistivity path where the work-function layers are no longer needed. In this field, the PMOS stacks leave room for W between ribbons, while the NMOS stacks occupy more of those gaps. A silicon-based dielectric marks the P/N boundary, allowing the PMOS and NMOS gates, sharing the same gate trench, to be processed sequentially. Fast logic paths, retention circuits, and SRAM need a family of threshold options. Changing threshold without substantially changing device dimensions, capacitance or fabrication complexity is valuable. FinFET processes typically use different work-function-metal stacks. In a four-ribbon GAA stack, the narrow sheet-to-sheet gap limits how much WFM can fit around each channel. La in the gate dielectric creates interfacial dipoles at the SiOx/HfOx boundary, shifting effective work function and tuning threshold voltage. This gives Intel another control alongside its NMOS and PMOS WFM stacks. Low-threshold devices improve critical-path drive; higher thresholds reduce leakage elsewhere. Dipole tuning is especially useful in GAA because it changes threshold without consuming the narrow intersheet gap with thicker WFM. Precise control of La incorporation, diffusion and interface quality has long been a challenge, limiting viability in high volume production but is now seen from every leading-edge foundry. Intel’s patent describes depositing a dipole-forming oxide above HfOx and annealing it toward the interfacial oxide before completing the work-function and fill metals. This separates threshold tuning from the space available for metal. Newer research addresses the thermal cost: imec’s 2026 dipole-middle research inserts the shifter between two HfOx depositions, shortening the diffusion path while protecting SiOx during patterning. [20], [21]Matched-cut EDS, Intel 18A (left) vs. Samsung SF2 (right). Intel retains raised source/drain epi beneath its contacts, while Samsung recesses W deep into the epi to form a V-shaped Ti-lined interface. The deeper contact increases metal-to-semiconductor area and shortens the current path from the lower sheets, reducing contact and spreading resistance. It also removes epi volume and brings the contact etch closer to the channel ends. Retaining more epi preserves the material available for strain transfer, especially from SiGe into PMOS. These geometries balance contact access against stress engineering and etch margin. [22], [23] Samsung stacks three sheets to Intel’s four ribbons, and both processes use sheet width to tune drive strength. In our Samsung cross-sections, widths range roughly from 19 to 30 nm in the NPU rows and 37 to 50 nm in the CU cell. The Samsung nanosheets taper, with the widest sheet at the bottom and the narrowest at the top. Both processes use HfOx gate dielectric and Ti-based work-function stacks, with Al in the NMOS stack. In the Samsung devices shown here, the dielectric and WFM occupy the intersheet gaps, leaving W above the top sheet. Intel’s PMOS stack leaves more room between ribbons, and W fills those gaps while the thicker NMOS stack leaves W mainly in the upper trench. Gate-stack EDS maps. The W between Intel’s PMOS ribbons provides a conductive path close to the lower gates. Where WFM fills the entire gap, the gate still surrounds the channel, but voltage reaches it through the more resistive work-function films. Thinner WFM and dipole tuning preserve room for low-resistivity fill; Mo and Ru are alternative fill metals being developed for further scaling. [24] A masked, sequential WFM flow explains the different gate heights and inter-nanosheet fill. The proposed sequence below shows how separate NMOS and PMOS work-function steps produce that geometry. Enabled by the BSPDN process, Intel replaces the dense-logic silicon subfin with dielectric, removing the parasitic conduction path below the ribbons and reducing substrate-related capacitance. A retained silicon body as in classical, non-SOI, planar and FinFET designs needs junction and punchthrough-stop engineering to suppress leakage. Dielectric isolation makes that leakage less sensitive to the subfin doping profile but adds removal and fill steps. It also weakens the direct thermal path through silicon, making the contacts, metal stacks and package more important for heat extraction. [24], [26] Fluorine is concentrated around selected Intel device structures in the maps. WF6 is a standard precursor for W nucleation and fill, while barrier films protect adjacent dielectrics from fluorine attack. Low-fluorine W processes reduce the residual-F burden. Chloride-based precursors avoid introducing F during W deposition, but require control of chlorine attack, nucleation and fill quality. The integration target is a continuous, low-resistance W path with a thin protective liner and minimal chemical damage to the surrounding stack. [20], [27], [28] The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page. We measured cell height, gate pitch, metal geometry, and ribbon dimensions at the XTEM sites shown below. The tables group these dimensions by site and device polarity. Our “sheet cuts” cross the silicon channel and show the ribbons end-on. “gate cuts” run along the channel through successive gates. The 18A logic cell dimensions point to a five-track logic library while the N3E and Intel 3 cell dimensions evidence a seven-track logic library. The DDR-PHY uses wider M0 wires and much larger spacing than core logic. That trades routing density for lower wire resistance and weaker coupling between neighboring nets. The geometry suits the current delivery and coupling requirements of analog, clock, and I/O circuitry. PowerVia lets 18A combine a compact cell height with wider M0 geometry by moving the main power rails off the signal-routing tracks. That relaxes local wire scaling while preserving a small cell footprint. Cell height and gate pitch set the geometric density; pin access and routability determine how much of it a real block can use. [29] The biggest takeaway from our gate-pitch measurements is that Intel 18A compute logic and TSMC N3E GPU logic have similar density in the Bohr representative-cell model. The 18A example is 18.6% denser than the Intel 3 GPU example. Gate pitches are nearly identical across the three sites, so cell height drives most of the difference. The Bohr model combines a four-transistor NAND2 spanning three gate pitches and a 32-transistor scan flip-flop (SFF) spanning nineteen pitches, weighting their densities 60:40. The sensitivity column shows how independently changing cell height and gate pitch by ±1 nm changes the result. This compares representative cell geometries; whole-die density also depends on cell mix and placement. The 18A P-core gives M0 substantially more metal cross section than the N3E vector engine. Treating each profile as a trapezoid gives 2.63 times the area per line and 1.84 times the area after normalization by routing pitch. The larger section reduces the geometric contribution to line resistance and lowers current density for a given current. Taller and wider wires also add capacitance, so circuit delay depends on the balance of resistance and capacitance. The DDR-PHY has less metal area per routing width than the 18A core fields, while remaining above N3E. [30] Area = height × (top CD + bottom CD) / 2, including liners. Area/pitch normalizes by routing width. Taper is the symmetric sidewall angle from vertical, with the largest angle belonging to the DDR-PHY. Compute tile The measurements show how ribbon dimensions and gate-stack geometry vary across the compute tile and between NMOS and PMOS to balance channel drive, gate load and the space needed for the dielectric/WFM stack across logic, SRAM and the DDR-PHY. Width mainly changes available channel perimeter; thickness also changes electrostatic control and carrier confinement. Gate-stack thickness then determines the space left for low-resistivity fill P-core and LP E-core logic Both the P-core and LP E-core use multiple nanosheet widths. Widths are measured on high-magnification XTEMs while wider-field images demonstrate additional width choices within the LP E-core. Multiple widths are expected even within an LP E-core. Timing-critical paths, buffers and cells with different fanout need different drive strengths. The lower-magnification fields show this width diversity beyond the sites quantified in the table. L2 and L3 SRAM GAA gives SRAM designers another way to balance the pull-up (PU), pass-gate (PG), and pull-down (PD) transistors. FinFET bitcells set device strength through fin count while GAA adds nanosheet width as a sizing knob. In a 6T SRAM cell, a strong pull-down relative to the pass-gate limits read disturbance, while a strong pass-gate relative to the pull-up improves writability. During a write, the pass-gate and write driver pull the node storing “1” below the inverter trip point. During a read, the pull-down holds the node storing “0” low. Bias, threshold voltage, mismatch and assist circuitry set the remaining margin. FinFET high-current cells commonly use a PU:PG:PD fin-count pattern of 1:2:2, a device-sizing ratio rather than a current ratio. Ribbon width lets Intel balance SRAM strengths without adding whole fins. The L2 cell uses its narrowest ribbons for PU and widest for PD, improving writability and read stability respectively. Intel’s disclosed HCC operates without assist; its denser HDC uses negative-bitline write assist. Pulling the selected bitline briefly below ground increases pass-gate overdrive so it can overpower the pull-up at lower supply voltage. That buys density and low voltage writability at the cost of boosting circuitry, switching energy, and additional voltage stress that must be controlled. [31], [32] Four rectangular ribbons give the perimeter = 8 × (width + thickness), before corner rounding. PG/PU is 1.49 and PD/PG is 1.16. The L3 structures closely resemble L2 in layout and cell height. Fewer L3 nanosheet widths are tabulated because fewer high-magnification images were available. DDR PHY The DDR-PHY trades density for controlled analog behavior and reliable off-chip signaling. It contains drivers, receivers, delay circuits, and calibration logic that set drive strength, sampling time, and voltage margin. Repeated four-sheet devices with similar widths fit the use of regular transistor units for matching and programmable drive. Its wider local wiring provides room for current delivery and separation of sensitive signals, while consuming more area than a dense core-logic grid. The layout serves the memory channel’s electrical requirements as well as digital logic density. [33] The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page. Intel 3 GPU devices Vector engine logic Intel 3’s XVE logic uses two-fin PMOS and NMOS devices with power rails in M0. Its cell height and M0 pitch give a seven-track geometry, two tracks more than the 18A logic. One-fin groups also appear among the two-fin devices. Intel 3 L2 SRAM The Intel 3 L2 SRAM uses the familiar HCC sizing pattern: one PU fin, two PG fins, and two PD fins. N3E GPU devices Vector engine logic The N3E XVE field contains repeated two-fin devices with seven-track cell geometry. N3E remains a FinFET process, giving Panther Lake a direct FinFET-to-RibbonFET comparison. N3E L2 SRAM The N3E L2 SRAM uses the same PU:PG:PD fin-count pattern of 1:2:2. Panther Lake-U follows Lunar Lake’s floorplan quite closely. Both pair 4 P-cores with 4 LP E-cores and NPU, media and display engines in similar locations. Lunar Lake also uses Xe2, the direct predecessor to Panther Lake’s Xe3 GPU. This makes Lunar Lake the most direct basis for our comparisons. Arrow Lake differs in core count and uses the older Xe-LPG GPU architecture, so we only use it where it offers a more direct component-level comparison. Compute tile Panther Lake compute-tile floorplans remain sparse even months after launch. Intel 18A’s backside metal and dielectric stack must be removed without damaging the underlying structures before a clean transistor-level floorplan can be imaged. Most published die shots hide or heavily process the background, but we are quite proud of the die shot we achieved and are excited to show the work we have done. We measured the areas of the key components on the compute tile and compared them with their Lunar Lake predecessors on TSMC N3B. These help us to capture changes in block area and compare the two chips across process nodes and designs. Our total tile areas exclude the scribe-line area. The compute-plus-GPU subtotal below uses the PTL-U compute tile and GT1 GPU; it excludes the I/O tile and passive base. Individual block areas use the boundaries marked on the floorplans The compute-plus-GPU row is recomputed from the displayed PTL-U and GT1 areas. Component rows use their stated per-region counts and are not an additive partition of the whole tile. The P-core area remains almost unchanged between Lunar Lake and Panther Lake, despite L2 capacity increasing from 2.5 MiB to 3 MiB. Arrow Lake uses the same Lion Cove core as Lunar Lake but also has a 3 MiB L2. Cougar Cove fits 20% more L2 into the same P-core area. The larger private cache keeps more of each core’s working set close to its execution units, reducing access to shared L3 and DRAM. Extra capacity adds storage leakage and lookup energy, so designers balance it against avoided lower-level accesses. The shared P-core L3 cache also shrank by 14.8%. [2] Cougar Cove combines a similar footprint with Intel’s reported power-efficiency improvements. RibbonFET’s tighter channel control reduces leakage, while PowerVia reduces supply droop and allows tighter voltage guardbands. [1] Darkmont’s four-core LP E-core cluster is 5.0% smaller than Skymont’s on Lunar Lake, with most of the reduction in its L2 regions. The 1 MiB region shrank by 8.4% and the 1.5 MiB region by 14.9%. The tag arrays also use one fewer visible row. Tags identify which memory addresses the data array holds, so rearranging them changes the cache’s layout and wiring without requiring less data capacity. [2] The LP E-cores share one L2. This pools capacity and avoids duplicating all the cache machinery, but the four cores contend for its banks and bandwidth. Their separate cluster also keeps light work away from the performance cluster and its L3, allowing that larger domain to sleep. [1], [2] Cache area includes more than the storage cells. Tags identify each line, decoders select rows, sense amplifiers read the small bitline signal, and wires connect to the banks. Splitting an array into smaller sections shortens wordlines and bitlines, improving access speed, but duplicates peripheral circuits. Panther Lake’s smaller cache regions therefore reflect the complete memory implementation, including how much of each region is devoted to storage. [34] Unlike Meteor Lake and Arrow Lake, Panther Lake has no separate SoC tile. The NPU, LP E-cores, memory controllers, PHYs, media and display engines now share the compute tile. This removes an active die and keeps CPU memory traffic on one die. The cost is moving PHY and I/O-related circuitry onto 18A: drivers, receivers and analog circuits must still meet external voltage, loading and signal-integrity requirements, so their area does not shrink like dense digital logic. [1], [2] The biggest shrink comes from the NPU, which occupies 36.9% less area. NPU 5 consolidates the same total INT8 MAC count into half as many neural compute engines. Each of the three NCEs has a larger MAC array to make the complete NCE envelope 22.6% larger than an NPU 4 engine. Consolidation also halves the number of scratchpads and SHAVE DSPs, from 12 to 6. The MAC array handles matrix multiplication and convolution, while SHAVE executes vector and custom operations that fit the array poorly. [1], [2], [35] The paired floorplans identify each NCE envelope and its scratchpad, MAC, and SHAVE regions. Each measured MAC polygon is counted once per NCE in the area accounting below. The scratchpads store weights, activations, and intermediate results near the MAC arrays, allowing repeated use without fetching them again from DRAM. Halving their number delivers the largest measured area saving but leaves less local storage for the same total MAC count. Layers that no longer fit locally require smaller working tiles or more transfers of intermediate data. The benefit depends on keeping the enlarged arrays busy while managing that tighter storage budget. [36] NPU 5 also adds native FP8. Using half the operand width of FP16 reduces storage and transfer demand, helping workloads fit the smaller local memory budget. Lower precision and format-dependent range make scaling and model validation part of deployment. Hardware activation functions further reduce work that would otherwise occupy the programmable DSPs. [1], [2] Microsoft requires an NPU to deliver at least 40 TOPS for Copilot+ PCs. Both Lunar Lake and Panther Lake meet this threshold, but Panther Lake uses significantly less silicon. GPU tiles Panther Lake is Intel’s first product with Xe3, its latest GPU architecture. It offers two different GPU tiles: a smaller GT1 tile with 4 Xe3 cores on Intel 3 and a larger GT2 tile with 12 Xe3 cores on TSMC N3E. Panther Lake allows us to compare the same GPU architecture across both Intel 3 and TSMC N3E. Wildcat Lake adds a third Xe3 implementation on Intel 18A. A future newsletter will detail Xe3 and its implementation differences across all three process nodes. GT2 scales Xe3 to a different physical layout, with render slices arranged vertically instead of GT1’s horizontal arrangement. Slice placement sets the distances to shared cache banks and the D2D interface. Those wires consume area and add delay, so scaling the number of Xe cores also requires a new balance of cache placement, routing and timing. [1] What’s immediately obvious is that the GT2 tile on TSMC N3E has much smaller Xe cores than GT1. These block areas include logic, caches, and routing. An Xe core on the GT1 tile is ~69% larger than one on Lunar Lake, and ~55% larger than one on GT2. Intel 3 therefore uses substantially more area per Xe core. The block-area gap exceeds the measured logic and SRAM density gaps, bringing routing, timing targets, cell mix, and floorplan allocation into the comparison. The measured vector/matrix engine region is almost unchanged between Lunar Lake and Panther Lake’s GT2 tile. Xe3 retains eight 512-bit vector engines and eight 2048-bit XMX engines per core. Its gains also come from feeding those engines more effectively: more resident threads hide stalls, and variable register allocation lets shaders trade registers per thread against the number of threads kept active. [1] The shared L1/SLM capacity increased by 33% from 192 KiB to 256 KiB, while its area increased only 5%, raising effective density by 27%. L1 retains reused cache lines, while software-managed SLM lets a thread group share data locally. Both reduce traffic to more distant memory. Allocating more SLM per group can also limit how many groups reside on a core at once. [1], [37] The GT1 tile carries 4 MiB of L2 against 16 MiB on the GT2 tile. GT1 divides its L2 cache into four 1 MiB banks, while GT2 uses eight 2 MiB banks. Each bank contains 128 macros, but each N3E macro stores 16 KiB, twice the Intel 3 macro’s 8 KiB capacity. The N3E macro is only 54% larger while holding twice as many bits, giving it 30% higher density: ~23.7 Mbit/mm² versus 18.3 Mbit/mm². Including bank-level circuitry, the gap widens to ~16.9 Mbit/mm² on GT2 versus ~10.4 Mbit/mm² on GT1. GT2 gains density with its macros storing more bits per unit area, and those macros occupy more of each cache bank. Larger macros spread decoder and sense-amplifier overhead across more storage, while a more compact bank layout reduces the share spent on control and routing. The compromise is longer wordlines and bitlines that carry more capacitance. [34] I/O tile Panther Lake uses two I/O tile variants, both fabricated on TSMC N6. The smaller one provides 4 PCIe 5.0 and 8 PCIe 4.0 lanes and serves lower-tier systems as well as those without a discrete GPU, while the larger one adds 8 PCIe 5.0 lanes, bringing the total to 20 lanes, for discrete-GPU connectivity. Panther Lake SKUs with the larger 10- or 12-Xe GPUs use the smaller I/O tile. [38] The smaller I/O tile adds a PCIe 4.0 block and a Thunderbolt block to Lunar Lake’s I/O layout, providing four additional PCIe 4.0 lanes and another Thunderbolt 4 port. Its repeated N6 blocks retain nearly identical areas and layouts. Reusing these proven PHYs and controllers avoids porting and requalifying external interfaces on 18A, where faster digital logic offers less benefit to circuits constrained by the off-chip link. [38] SemiAnalysis’s teardown lab (STEEL) dives deep into the world’s advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. We’re hiring technical experts from system to transistor and everywhere in between. Check out our Careers page. Panther Lake offers scalability and modularity through its disaggregated packaging that partition compute, GPU, and I/O silicon into separate tiles allowing for a suite of tile configurations. This partitioning makes the package part of Intel’s node economics as it determines how much leading-edge wafer area each product consumes, which functions can remain on other processes, and how much configuration freedom Intel can offer from a shared set of tiles. Furthermore, fabricating the compute and GPU tiles separately confines the new 18A process to the compute tile and allows graphics and I/O to use other, more established, and more cost-effective processes. For Panther Lake, the GPU and I/O tiles are assembled alongside the compute tile on a passive silicon base using Foveros-S. Intel’s current technology brief lists a nominal 36 µm pitch for Foveros-S. Through-silicon vias (TSVs) in the base connect the fine wiring above to the larger package connections below. The functional tiles sit side by side on that passive base in a 2.5D configuration. [39] Our cross-section through the compute and GPU tiles shows the package’s wiring hierarchy. Microbumps connect each active tile to the passive silicon base; its fine redistribution layer (RDL) carries the short, dense tile-to-tile links. TSVs carry connections through the base to the package substrate, which fans them out to the much coarser motherboard solder joints. The base supplies interconnect, while computation remains in the active tiles above it. [39] At the compute-tile edge, the higher-magnification inset shows a local microbump spacing of approximately 25.24 µm and a feature width of 12.33 µm. These local spacings are finer than Intel’s nominal Foveros-S value. The X-ray fields further confirm tighter neighboring bumps, consistent across every die-to-die area found on each tile. Additional X-ray analysis is offered after the paywall. Putting the memory controller beside the CPU removes the D2D transfer that CPU memory requests required in Meteor Lake and Arrow Lake. This avoids the extra transmitter, receiver, and link traversal, saving interface energy and latency. Panther Lake’s separate GPU still crosses a D2D link to reach DRAM, so its larger local caches also help contain package traffic. [1], [40] Smaller dies are less likely to contain a random fatal defect, and screening them before assembly prevents one bad tile from consuming a complete package of good silicon. Reuse also spreads design and qualification work across more products. Against those gains, Intel pays for the passive base, D2D circuits, extra bonding and test steps, and losses during assembly. Cost per working product across the portfolio captures the combined effect of wafer yield, reuse, test, and assembly. [29] Wildcat Lake packaging Intel launched Core Series 3, formerly Wildcat Lake, on 16 April 2026 for value mobile and edge systems. Wildcat Lake keeps 18A but removes the passive base and combines more functions on one die to simplify the package. The two products therefore reveal two distinct ways to commercialize the same leading-edge process. [41] Wildcat Lake’s 18A die combines up to two Cougar Cove P-cores, four Darkmont LP E-cores, two Xe3 cores and a smaller NPU. A separate platform-controller die supplies I/O, connected through UCIe, Intel’s first processor implementation of the standard. Consolidating graphics remove a tile boundary and the passive base, reducing assembly complexity for a modest-bandwidth value product. It also ties CPU and graphics scaling to the same die, giving up Panther Lake’s ability to swap in a much larger GPU. [42], [43] In July 2021, Intel CEO Pat Gelsinger set out an ambitious process roadmap aimed at regaining performance leadership by 2025, later described as five nodes in four years. Five years and one CEO later, Intel’s comeback story is not as unambiguously positive as Pat may have hoped. [44], [45] Intel once set the pace for process technology, bringing high-k metal gate technology and FinFETs into volume production years ahead of the rest of the industry. Its 22 nm FinFET process reached consumers with Ivy Bridge in 2012. [46] Intel’s integrated device manufacturing (IDM) model allowed its architects and process engineers to co-optimize products and processes. Starting with Sandy Bridge, Intel dominated x86, while AMD struggled with Bulldozer. That lead faltered at 14 nm and broke at 10 nm. Intel targeted a massive 2.7× density increase, but the node arrived years late and required several revisions before it could support Intel’s full lineup. This delay forced Intel to stretch 14 nm across six generations, while TSMC moved ahead in process technology and AMD recovered in x86. By 2019, Intel was still shipping 14 nm across most of its product stack, with its 10 nm client ramp focused on Ice Lake mobile processors. Meanwhile, TSMC was shipping N7 and N7+, and AMD’s Zen 2 compute chiplets used N7 to raise core counts and improve efficiency. Intel’s process failures were central to its decline, but unsound business decisions furthered their downward slide. Product delays compounded product mistakes, pushing client, server, and FPGA roadmaps off schedule. Several attempts to enter AI (Nervana and Gaudi) and networking (Tofino) also failed to establish lasting businesses. Intel’s recovery has focused on consumer CPUs and advanced packaging. Tiger Lake, Alder Lake, Lunar Lake and now Panther Lake have restored Intel’s consumer roadmap. On the process side, Intel 4 shipped with Meteor Lake, Intel 3 with Granite Rapids and Sierra Forest, and Intel 18A with Panther Lake. Intel has also made advanced packaging part of its foundry offering. However, Intel is still playing catch-up in servers. Several Xeon generations arrived years late and trailed contemporary AMD and Arm server CPUs in performance, efficiency, and core count. The process roadmap is back, but Intel does not hold the same process-technology leadership position it held prior to 10 nm. The introduction of gate-all-around nanosheets and backside power delivery are two of the biggest changes to transistor integration in a decade. Intel took on both changes at once: 18A paired its first RibbonFET with PowerVia in Panther Lake. Panther Lake is a substantial manufacturing milestone. Our cross-sections show how RibbonFET and PowerVia reshape local contacts and wiring, while the floorplans show where architectural consolidation and process choices save area. A sustained competitive lead depends on product performance, cost, yield, and the next implementation. The SemiAnalysis STEEL teardown lab breaks down advanced datacenter and AI hardware. To learn more about our pipeline or to commission a teardown, contact sales@semianalysis.com. WE’RE HIRING: Architecture, floorplan, packaging, manufacturing, and labs experts. Opportunities from system to transistor and everywhere in between. Check out our Careers page.. [1] Intel, “Intel Tech Tour 2025 Panther Lake recap,” Doc. 866361, 2025. Accessed: Sep. 14, 2026. [Online]. Available: https://cdrdv2-public.intel.com/866361/ITT_2025_Panther_Lake_Recap1.pdf [2] Intel, “Core Ultra processors Series 3 datasheet, volume 1,” Doc. 872188-001, Jan. 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://cdrdv2-public.intel.com/872188/872188-001.pdf [3] N. Horiguchi and E. Beyne, “Backside power delivery: How to power chips from the backside,” imec Reading Room, Nov. 25, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://www.imec-int.com/en/articles/how-power-chips-backside [4] W. Hafez et al., “Intel PowerVia technology: Backside power delivery for high density and high-performance computing,” in Proc. IEEE Symp. VLSI Technol. Circuits, Kyoto, Japan, Jun. 2023, pp. 1–2, doi: 10.23919/vlsitechnologyandcir57934.2023.10185208. [5] K. Fischer et al., “Intel 18A platform technology featuring RibbonFET (GAA) and PowerVia for advanced high-performance computing,” in Proc. Symp. VLSI Technol. Circuits, Kyoto, Japan, Jun. 2025, pp. 1–3, doi: 10.23919/vlsitechnologyandcir65189.2025.11075006. [6] K.-F. Cheng, C.-L. Teng, H.-Y. Huang, and H.-C. Chen, “Metal oxide composite as etch stop layer,” U.S. Patent 12 176 247 B2, Dec. 24, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US12176247B2/en [7] S. W. King, “Dielectric barrier, etch stop, and metal capping materials for state of the art and beyond metal interconnects,” ECS J. Solid State Sci. Technol., vol. 4, no. 1, pp. N3029–N3047, 2015, doi: 10.1149/2.0051501jss. [8] A. V. Mule’, D. J. Towner, D. Seghete, C. R. Ryder, and A. A. Gonzalez, “Multi-layer etch stop layers for advanced integrated circuit structure fabrication,” U.S. Patent Appl. 20220102343 A1, Mar. 31, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20220102343A1/en [9] C.-C. Wang and J. H. Wang, “Interconnect structure with dielectric cap layer and etch stop layer stack,” U.S. Patent Appl. 20230253247 A1, Aug. 10, 2023. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20230253247A1/en [10] Y. J. Wu, K.-F. Cheng, C.-C. Lee, H.-K. Chang, and H.-Y. Huang, “Etch stop layer for interconnect structures,” U.S. Patent Appl. 20240332070 A1, Oct. 3, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20240332070A1/en [11] M. G. Rainville, N. Shankar, K. S. Reddy, and D. M. Hausmann, “Deposition of aluminum oxide etch stop layers,” U.S. Patent Appl. 20180197770 A1, Jul. 12, 2018. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.justia.com/patent/20180197770 [12] J. S. Leib et al., “Conductive lines having molybdenum liner and tungsten fill for advanced integrated circuit structure fabrication,” U.S. Patent Appl. 20240429126 A1, Dec. 26, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.justia.com/patent/20240429126 [13] M. Naik, “Cobalt enables power and performance scaling at single-digit logic nodes,” Applied Materials, Dec. 17, 2018. Accessed: Sep. 14, 2026. [Online]. Available: https://www.appliedmaterials.com/us/en/blog/blog-posts/cobalt-enables-power-and-performance-scaling-at-single-digit-logic-nodes.html [14] Lam Research, “Breaking through AI’s invisible barrier with molybdenum,” Lam Research Newsroom, May 6, 2025. Accessed: Sep. 14, 2026. [Online]. Available: https://newsroom.lamresearch.com/molybdenum-metallization-ai-revolution [15] P. Yashar, G. Malyavanatham, and H. Vijwani, “Integrated circuit interconnect structures with niobium barrier materials,” U.S. Patent Appl. 20240112951 A1, Apr. 4, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20240112951A1/en [16] Applied Materials, “Applied Materials unveils chip wiring innovations for more energy-efficient computing,” Applied Materials Investor Relations, Jul. 8, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://ir.appliedmaterials.com/news-releases/news-release-details/applied-materials-unveils-chip-wiring-innovations-more-energy [17] Samsung, “Samsung showcases AI-era vision and latest foundry technologies at SFF 2024,” Samsung Semiconductor, Jun. 13, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://semiconductor.samsung.com/news-events/news/samsung-showcases-ai-era-vision-and-latest-foundry-technologies-at-sff-2024/ [18] C. W. Yeung et al., “Channel geometry impact and narrow sheet effect of stacked nanosheet,” in Proc. IEEE Int. Electron Devices Meeting (IEDM), Dec. 2018, pp. 28.6.1–28.6.4, doi: 10.1109/iedm.2018.8614608. [19] S. Mochizuki et al., “Evaluation of (110) versus (001) channel orientation for improved nFET/pFET device performance trade-off in gate-all-around nanosheet technology,” in Proc. Int. Electron Devices Meeting (IEDM), Dec. 2023, pp. 1–4, doi: 10.1109/iedm45741.2023.10413854. [20] D. G. Ouellette et al., “Fabrication of gate-all-around integrated circuit structures having molybdenum nitride metal gates and gate dielectrics with a dipole layer,” U.S. Patent 12 051 698 B2, Jul. 30, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US12051698B2/en [21] imec, “Advancing the CFET-based device roadmap with novel integration modules and standard-cell architectures,” imec, 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.imec-int.com/en/articles/advancing-cfet-based-device-roadmap-novel-integration-modules-and-standard-cell [22] A. Reznicek, X. Miao, C. Lee, and J. Zhang, “Wrap around contact for nanosheet source drain epitaxy,” U.S. Patent 11 302 813 B2, Apr. 12, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US11302813B2/en [23] J. A. Kittl, J. G. Hong, D. R. Palle, and M. S. Rodder, “Methods for varied strain on nano-scale field effect transistor devices,” U.S. Patent 9 978 833 B2, May 22, 2018. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US9978833B2/en [24] S. Gandikota et al., “Method of reducing metal gate resistance for next generation NMOS device application,” Int. Patent Appl. WO 2024/137272 A1, Jun. 27, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/WO2024137272A1/en [25] J. Zhang et al., “Full bottom dielectric isolation to enable stacked nanosheet transistor for low power and high performance applications,” in Proc. IEEE Int. Electron Devices Meeting (IEDM), Dec. 2019, pp. 11.6.1–11.6.4, doi: 10.1109/iedm19573.2019.8993490. [26] C. Yoo, J. Chang, Y. Seon, H. Kim, and J. Jeon, “Analysis of self-heating effects in multi-nanosheet FET considering bottom isolation and package options,” IEEE Trans. Electron Devices, vol. 69, no. 3, pp. 1524–1531, Mar. 2022, doi: 10.1109/ted.2022.3141327. [27] S. S. Pradhan, D. B. Bergstrom, J.-S. Chun, and J. Chiu, “Tungsten gates for non-planar transistors,” European Patent Appl. EP 3 506 367 A1, Jul. 3, 2019. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/EP3506367A1/en [28] L. Schloss and X. Ba, “Tungsten films having low fluorine content,” U.S. Patent 9 754 824 B2, Sep. 5, 2017. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US9754824B2/en [29] Intel Foundry, “Accelerating AI and HPC with advanced process and packaging technologies,” Doc. 367370-001US, Jun. 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/dam/www/central-libraries/us/en/documents/2025-11/intel-foundry-hpc-ai-brief.pdf [30] M. Lofrano, X. Chang, H. Oprins, and Z. Tokei, “Mitigating the thermal bottleneck in advanced interconnects,” imec, Sep. 28, 2023. Accessed: Sep. 14, 2026. [Online]. Available: https://www.imec-int.com/en/articles/mitigating-thermal-bottleneck-advanced-interconnects [31] S. Bair, “Intel at ISSCC 2025: Navid Shahriari invited talk, eight papers, forums, panelist & product details,” Intel Community, Feb. 19, 2025. Accessed: Sep. 14, 2026. [Online]. Available: https://community.intel.com/t5/Blogs/Tech-Innovation/Edge-5G/Intel-at-ISSCC-2025-Navid-Shahriari-Invited-Talk-Eight-Papers/post/1667592 [32] D. Chandra, E. Potladhurthi, D. R. S. Reddy, and K. S. Rengarajan, “Tunable negative bitline write assist and boost attenuation circuit,” U.S. Patent Appl. 20160203857 A1, Jul. 14, 2016. Accessed: Sep. 14, 2026. [Online]. Available: https://patents.google.com/patent/US20160203857A1/en [33] Synopsys, “Advantages of firmware-based training in high-speed DDR IP,” Synopsys IP Technical Bulletin. Accessed: Sep. 14, 2026. [Online]. Available: https://www.synopsys.com/articles/firmware-based-training-ddr-ip.html [34] N. Muralimanohar, R. Balasubramonian, and N. P. Jouppi, “CACTI 6.0: A tool to understand large caches,” Hewlett-Packard Laboratories, 2009. Accessed: Sep. 14, 2026. [Online]. Available: https://users.cs.utah.edu/~rajeev/cacti6/cacti6-tr.pdf [35] Intel, “Intel Tech Tour 2024 Lunar Lake AI hardware accelerators,” Doc. 824436, 2024. Accessed: Sep. 14, 2026. [Online]. Available: https://cdrdv2-public.intel.com/824436/2024_Intel_Tech%20Tour%20TW_Lunar%20Lake%20AI%20Hardware%20Accelerators.pdf [36] Y.-H. Chen, J. Emer, and V. Sze, “Eyeriss: A spatial architecture for energy-efficient dataflow for convolutional neural networks,” in Proc. 43rd ACM/IEEE Annu. Int. Symp. Comput. Archit. (ISCA), Jun. 2016, pp. 367–379, doi: 10.1109/isca.2016.40. [37] Intel, “Introduction to the Xe-HPG architecture,” Doc. 758306, Nov. 4, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/www/us/en/developer/articles/technical/introduction-to-the-xe-hpg-architecture.html [38] Intel, “Core Ultra processors Series 3 for the edge,” Doc. 855291. Accessed: Sep. 14, 2026. [Online]. Available: https://cdrdv2-public.intel.com/855291/Intel%C2%AE%20Core%E2%84%A2%20Ultra%20Processors%20Series%203%20for%20Edge%20Overview_2.pdf [39] Intel Foundry, “Foveros technology brief,” Doc. 366411-001US, Feb. 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/dam/www/central-libraries/us/en/documents/2025-07/foveros-25d-product-brief.pdf [40] D. Das Sharma, “UCIe: Building an open chiplet ecosystem,” UCIe Consortium, 2022. Accessed: Sep. 14, 2026. [Online]. Available: https://www.uciexpress.org/_files/ugd/0c1418_c5970a68ab214ffc97fab16d11581449.pdf [41] Intel, “Intel launches Intel Core Series 3 processors,” Intel Newsroom, Apr. 16, 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/www/us/en/newsroom/news/client-computing/intel-launches-intel-core-series-3-processors-changing-the-game-for-everyday-computing.html [42] Intel, “Core Series 3 launch press deck,” Apr. 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://download.intel.com/newsroom/2026/Intel-Core-Series-3/Intel-Core-Series-3-Launch-Press-Deck.pdf [43] Intel, “Intel outlines architectures for agentic AI at Hot Chips 2026,” Intel Newsroom, Aug. 24, 2026. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intel.com/content/www/us/en/newsroom/news/client-computing/intel-outlines-architectures-for-agentic-ai-at-hot-chips-2026.html [44] Intel, “Intel accelerates process and packaging innovations,” Intel Investor Relations, Jul. 26, 2021. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intc.com/filings-reports/all-sec-filings/content/0001193125-21-224438/d199788dex991.htm [45] Intel, “Intel reports third-quarter 2021 financial results,” Intel Investor Relations, Oct. 21, 2021. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intc.com/news-events/press-releases/detail/1505/intel-reports-third-quarter-2021-financial-results [46] Intel, “3rd generation Intel Core processors bring exciting new experiences and fun to the PC,” Intel Investor Relations, Apr. 23, 2012. Accessed: Sep. 14, 2026. [Online]. Available: https://www.intc.com/news-events/press-releases/detail/577/3rd-generation-intel-core-processors-bring-exciting
CyberPowerPC Gaming PCs With Founder's Edition Nvidia GeForce RTX 5070 GPUs Are on Sale During the Walmart GeForce Week Sale Event
Part of the Walmart's GeForce Week Sale that ends on September 27.
CyberPowerPC Launches RTX Founders Edition Pre-Built PCs Starting at $1,549
NVIDIA's Founders Edition graphics cards are generally restricted to its own retail channels, but it seems as though CyberPowerPC has a limited run of RTX 50-series GPUs available in its pre-built gaming PCs for a GeForce Week event, which is slated to run until September 27. The CyberPowerPCs...
CyberPowerPC now sells prebuilt PCs with NVIDIA RTX 50 Founders Edition GPUs, 1,000 RTX 5080 systems sell in one day
CyberPowerPC is using NVIDIA's own RTX 50 Founders Edition cards in new Walmart systems. The RTX 5080 configuration with Ryzen 7 9800X3D costs $2,299.
GPUS Stock Slips After Hyperscale Data Carves Out Bitcoin Mining Business Into New Unit
Hyperscale Data management said Bitcoin mining and AI infrastructure require different capital strategies and priorities.
World of Warcraft: Forever drops GeForce GTX 780 support, but 2012 Radeon HD 7850 is still supported
World of Warcraft: Forever raises its official GPU minimum to GTX 10 series and RDNA 1. A separate architecture cutoff still allows some much older GPUs.
Engrams Embedding Entendre: Codesign for Efficient DRAM/SSD Offloading
Engram extends standard token embeddings with learned multi-token lookups. Recurring local patterns retrieve vectors directly, reducing the need to reconstruct them through attention and feed-forward layers. With Engram model architecture optimization, it allows for lower HBM capacity to be needed for models at the same quality. This does not mean there won’t be an insane demand for HBM but it just means that model architecture will continue to innovate around constraints. This model architecture design is naturally codesigned for parameter offloading: each token accesses a few embedding rows whose addresses depend on token IDs, not hidden states. The runtime can prefetch those rows from host DRAM while earlier layers compute, keeping the table outside HBM without transferring entire weight matrices. Our Memory model contains our latest estimates of quarter by quarter HBM, DRAM, & NAND supply and demand. Offloading frees HBM for model weights and KV cache, potentially supporting larger batches or more concurrent sessions. When DRAM becomes the next constraint, NVMe offers another tier. Recommendation systems already cache frequently or recently accessed embedding rows in faster memory while backing colder rows with SSDs. After NVIDIA roadmap had to change due to massively despec’ing Rubin Ultra from 1024GB to now ~200GB of HBM per chip, model architecture optimizations like emgram maybe helpful. Our DeepSeek-V4.1-Flash configuration uses roughly 189 GiB of memory for Engram. We replace it with a memory-mapped (mmap) file and measure serving performance with offload to SSD. Later on in our report, we will show our Engram offloading experiments along with the official InferenceX agentic inference serving results on Engram models like DeepSeekv4.1 Flash across all 6 NVIDIA GPU SKUs along with MI355X. Unsurprisingly, the CUDA Moat is still mogging MI355X on the ultra popular DeepSeekV4.1 Flash model. We also show how even on high capacity HBM SKUs, offloading emgrams to DRAM could result in even better performance for most of the pareto than keeping the emgram in HBM. Our benchmark has been widely reproduced, validated and/or supported by almost every major buyer of compute from Google Cloud to Microsoft Azure to Oracle, to Meta and many more. Furthermore, it has the support of the ML community including from vLLM, LMCache, SGLang, PyTorch, Huggingface and the support of major labs like OpenAI, MiniMax, ZAI, Qwen, Moonshot Kimi, etc. Star the InferenceX GitHub repository if you find the open-source benchmark and data useful!. InferenceX is the only inference benchmark in the world to have TPUv7, Jalapeño, Nvidia Rubin NVL72, AMD, and soon, SambaNova and Trainium. Due to how realistic AgentX scenario is to real world agentic inference workloads, AMD has committed to collaborating on MI455X UALoE72 too. DeepSeek did not release the original paper’s two trained Engram models. We replicated its setup on fineweb-edu using the released code and training hyperparameters, at an estimated 6E18 FLOPs per run. We observed the same U-shape scaling: Engram improved performance over pure MoE baselines. We also reproduced DeepSeek’s results, where earlier-layer representations with Engram resembled those of later layers. Like the original Engram paper, it is possible to probe Engram’s gate scores to see what n-grams DeepSeek-V4.1-Flash makes the most use of. Our gate scan finds names, code fragments, relational phrasing, and boilerplate. These examples prioritize interesting-ness over gate strength. One unexpected result was Wright : Ace Attorney. These examples suggest learned memory optimizes the training objective, not a judgment of which facts deserve storage. Licenses, bibliography fragments, API scaffolding, and website furniture can provide prediction shortcuts, so the value of additional Engram capacity may depend on what survives data preparation. This does not show that table capacity is “wasted”: the evaluation-corpus scan establishes neither training exposure nor the capacity occupied by each category. For offloading, strong gates do not identify cache-hot rows. Low gates do not automatically save reads either: computing the gate requires the retrieved key and undoes the performance gain of a fused kernel. Skipping reads would require a separate usefulness predictor before retrieval. In the original paper’s inference-time ablation, factual-knowledge benchmarks retained just 29–44% of their original performance, while reading comprehension retained 81–93%. This is due to the training–inference mismatch. The resulting degradation therefore measures this trained model’s dependence on Engram, not the performance difference between models trained with and without it. In our ablations, suppressing Engram worsens token likelihood across all evaluated domains, especially encyclopedia text and several code corpora. Surprisingly, GSM8K accuracy stays within measured run-to-run variation and removing Engram has no effect. Engram is not a detachable dictionary beside an unchanged MoE. Removing it changes downstream features and expert selection. We tested whether rerouting hurts or compensates by holding tokens fixed in a teacher-forced experiment on CRUXEval, a code-reasoning benchmark of small Python functions where the model predicts a function’s output from its code and an input, and scoring the reference answer. Removing Engram raised answer loss from 0.2848 to 0.3093 bits/token. Forcing the ablated model to use the original Engram-on expert choices made it worse still, at 0.3375 bits/token. Rerouting partially compensates for the missing memory. Memory features and expert selection work together, rather than following a clean “memory stores facts; experts reason” division. On the same CRUXeval, removing Engram during either phase reduced accuracy and increased generated tokens; removing it throughout produced the largest changes. Keeping Engram for prefill leads to more correct answers than keeping for decode likely due to semantically richer KV cache transferred to decode workers, allowing it to mitigate some of the performance loss. Engram’s table is large, but each lookup is small. DeepSeek-V4.1-Flash requests 24 rows at each of two Engram layers, about 12.4 KiB per processed token position across the model, or 3.1 KiB per GPU when split across four GPUs. Currently as of Day 7 since Model Release, MI355X is still 2-4x worse performance per dollar compared to B200 even when normalized by Mi355X’s lower TCO. Our full total cost of ownership breakdown comes from our AI Cloud TCO Model along with monthly market surveys of over 100+ gpu clouds & gpu cloud customers. On the Day 0 release of DeepSeekv4.1 Flash, NVIDIA vLLM works out of the box with zero issues across all 6 SKUs: H100, H200, B200, B300, GB200, GB300! This was thanks to the amazing work by the NVIDIA & Interact teams! In comparison, AMD vLLM did not work on day 0 for DeepSeekv4.1 Flash.Source: SemiAnalysis InferenceX AMD’s vLLM documentation points to using vllm/vllm-openai-rocm:deepseekv41-flash-0909, but from hour 0 of the model release to hour 23, AMD has not publicly released the image. AMD claims, “SPEED IS THE MOAT,” yet it has still not released it by the 23rd hour. We wish that, going forward, the AMD team has a better process for hour 0 model releases. Eventually when they did publicly release for “day 0” image support, performance-wise, it is currently up to 14.8x worse perf per dollar than H200 and up to 42x worse perf per dollar than B200/B300. The power of the CUDA MOAT is NVIDIA’s collaboration with its massive 6 million-developer community ecosystem including most of the vLLM & SGLang & Tokenspeed maintainers which means that CUDA is optimized on day 0. Overall, AMD did make significant improvements but the performance per dollar is still currently 2-4x worse than B200. AgentX Engram DRAM Offloading Improving Performance The HBM and DRAM offload use the same GPU kernel to select and dequantize rows. With HBM, it reads GPU memory; with Unified Virtual Addressing (UVA), it reads pinned host memory directly. Both support full decode graphs. Moving the table into HBM accelerates only the sparse lookup, leaving decoder computation and communication unchanged, resulting in little overall benefit while consuming memory otherwise available to KV cache. Another benefit of Engram offloaded to DRAM which means you can reduce the communication overhead by using less HBM GPUs per replica. For example, when enabling Engram offloading on B300, we are able to switch from TP4 to now TP2 which improves the pareto curve by up to 1.6x. When iso-model quality, less HBM is required as the engrams could be offloaded to host DRAM. Thus HBM bandwidth matters way more than HBM capacity. For inference workloads where memory bandwidth matters the most, 4-hi HBM provides the best $/bandwidth and therefore lowest cost per token. If China continues to make more and more revolutionary model architecture innovations, soon it could potentially 0Hi HBM stacks. [ ](https://newsletter.semianalysis.com/p/long-live-the-short-king-why-4-hi) [ Long Live the Short King: Why 4-hi HBM Wins ](https://newsletter.semianalysis.com/p/long-live-the-short-king-why-4-hi) Myron Xie, Bryan Shan, and 3 others · Sep 13 [ Read full story ](https://newsletter.semianalysis.com/p/long-live-the-short-king-why-4-hi) Moreover, on B300 and week-0 stack, moving the Engram table back to HBM did not improve results and stay within run-to-run variance. This is a result of the work optimizing DRAM offload, such as async, and overlap. SSD Offloading On B200, we created a unoptimized vLLM fork and stored the Engram tables in memory-mapped files on local SSDs. File backing lets the OS reclaim table pages when other applications need RAM. Pages already cached in memory can be served without reading the SSD again. Furthermore, note that we were unable to turn on GDS The unoptimized SSD implementation changes how rows reach the GPU. It copies row IDs to the CPU, deduplicates them, gathers the requested rows into pinned buffers, copies those rows back to the GPU and dequantizes them. This work runs between segments of the GPU execution graph. Native UVA performs row selection and dequantization directly on the GPU, avoiding the CPU round trip. A warm filesystem cache removes physical SSD reads, but leaves the coordination, row gathering and transfers. This is why a file already cached in RAM can still perform worse than a pinned DRAM table. The comparison measures the whole serving path; it does not separate the time spent on each of these operations. B200 DRAM dominates both measured SSD serving curves in total tokens per dollar and P90 interactivity. Near 125 tokens/s/user, DRAM delivers 121 million total tokens per dollar versus 52 million for SSD. For production serving, SSD offloading is likely not worth the tradeoff. On the B200 configurations we measured, SSD offloading loses on both measures: every observed SSD point has a DRAM alternative that delivers higher P90 interactivity and more total tokens per dollar. The only points where Cheaper storage does not automatically produce a cheaper inference service. Moving Engram to SSD leaves the same four expensive GPUs and the rest of the server in place. Reclaiming RAM only creates an economic benefit if it enables a cheaper server configuration or additional useful capacity. The current unoptimized path provides neither benefit, and the filesystem cache still consumes RAM when table pages are resident. Next we will look at the mechanisms and specific implementation of ngrams from DeepSeekv4.1 Flash, LongCat, Qwen3.8 Flash Next.
AMD Plans 10% Price Hike Across GPUs, Chipsets, and Possibly CPUs
AMD is preparing for a series of price hikes in response to rising costs of acquiring silicon from its supply chain. According to ChannelGate, AMD is working on a 10% price increase across its products due to the upstream supply chain squeeze. Since most of AMD's silicon is manufactured by...
Leaker Claims AMD And NVIDIA Delay Next-Gen GPUs To 2028, Except One RDNA 5-Based AT2 Chip Launching In 2027
AMD and NVIDIA may skip 2027 for next-gen GPUs, with only AMD's mid-range RDNA 5 AT2 launching next year alongside continued high prices.
DLSS Multi Frame Generation Has Now Been Modded Onto RTX 40, RTX 30 & Even RTX 20 GPUs
Modders unlock NVIDIA's DLSS Multi Frame Generation on RTX 20, 30, and 40 Series, expanding frame interpolation capabilities beyond the RTX 50 Series
Modders Unlock DLSS Multi Frame Generation For RTX 30-Series "Ampere" GPUs
After porting DLSS 5 to every recent NVIDIA generation, the modding community now seems to be turning its focus to Frame Generation. A new mod called DLSSG SM86 lets GeForce RTX 30-series owners run DLSS Multi Frame Generation in 2X and 4X modes, a feature NVIDIA officially reserves for RTX...
AMD Threadripper Halo Station packs 96-core CPU and Instinct MI350P GPUs, up to 576GB GPU memory and 2TB system memory
AMD has shown its Threadripper Halo Station with a 96-core CPU, two Instinct MI350P GPUs and 2TB of system memory. The design is planned to support four GPUs with 576GB HBM3E.
NVIDIA Confirms DLSS 5 Support Extends To RTX 40 GPUs, But First Focusing On Model Optimization on RTX 50 Series
NVIDIA DLSS 5 is going to be released on older RTX 40 GPUs and will be available once the model is fully tuned on RTX 50 series.
Nvidia will officially bring DLSS 5 to older GPUs — but won’t give gamers full control
The cat is out of the bag.
Nvidia wants to turn your house of gaming PCs into an AI supercomputer
Nvidia's free RTX PAIR app spreads local AI workloads across multiple GPUs on your network, dramatically cutting down on processing times.
DLSS 5 Officially Goes Live September 3 With NBA 2K27, Exclusive to RTX 50-series "Blackwell" GPUs
NVIDIA has confirmed that DLSS 5 will officially launch on September 3 alongside NBA 2K27, and it will remain exclusive to GeForce RTX 50-series GPUs. When asked directly at a Gamescom technical briefing whether older cards would ever receive support, NVIDIA gave a flat answer: "It's RTX...
NVIDIA DLSS 5 Performance Unveiled: 5x Faster Than Original Demo, 50% Perf Penalty on Average, But RTX 50 GPUs Offer “Hundreds of FPS” With MFG Enabled & More Optimizations On the Way
NVIDIA is sharing the first performance figures of its RTX 50 GPUs running DLSS 5, offering next-gen visuals at super-fast frame rates.
Leaked DLSS 5 is Being Tested on RTX 20-series "Turing" GPUs, Now Runs on Emulators and Old DirectX Titles
The leaked DLSS 5 DLL keeps finding new homes, and support has now spread to RTX 20-series "Turing" GPUs. That architecture is three whole generations behind the NVIDIA-recommended "Blackwell," and modders are showing no signs of stopping either. Developer ShortFuse has released new DLSS-NR...
Xbox CEO Reacts To The Irony Of Rising Console Prices Being An Issue That Microsoft Itself Contributed To
Xbox CEO Asha Sharma believes it is "wonderful" that GPUs have use beyond gaming.
NVIDIA Shuts Down RTX 5000 Power Limit Control OC in Latest Driver Update
We recently reported on the open-source overclocking tool, mVolt+, unlocking power limit control on NVIDIA GeForce RTX 5000-series GPUs, allowing certain GPUs, like the RTX 5090, to draw up to 125 W of additional power and enabling significantly higher core clock speeds as a result. With the...
DLSS 5 Patched to Work on RTX 4000 "Ada Lovelace" GPUs Despite No Official Support
The modding community has been hard at work ever since the untimely release of DLSS-NR, a precursor to DLSS 5. Now, modders have gotten NVIDIA's leaked DLSS 5 tool working on RTX 4000 series cards, even though the tool was only built for RTX 5000 Blackwell hardware. The DLL leaked earlier this...
Nvidia earnings could reveal its fastest-growing business isn't chips: Chart of the Day
Wall Street is focused on Nvidia's GPUs, but the faster-growing story is sitting beside them.
Intel Crescent Island GPUs Pack Up To 32 Xe3P Cores, Optimized For Agentic AI With Low-Cost LPDDR5X That Reaches Up To 480 GB Capacity
Intel has given the rundown on its next-gen AI inference accelerator, Crescent Island, which packs 32 Xe3 cores and 480 GB of LPDDR5X memory.
The Reason Why Data Center GPUs Have Such Short Lifespans
Because they're in constant use, become exposed to high heat levels, and rarely undergo maintenance, data centers run through GPUs in just one to three years.
NVIDIA RTX 50-Series GPUs Could See Another 20-30% Price Hike in 2026
We've previously reported on expected price increases for NVIDIA GPU kits that the company sells to OEMs, thanks to the AI boom and its resulting impact on the DRAM supply chain. Now, thanks to a report by Taiwanese publication, Economic Daily Times, it's been revealed that NVIDIA's price hike...
Nvidia GPUs see yet another price hike of up to 30 percent, as the generative AI hardware crisis worsens
In response Samsung announced another price increase across DRAM as the personal computing nightmare deepens.
Sapphire Intros Radeon RX 9050 8GB Pulse Graphics Card
Sapphire today introduced its Radeon RX 9050 Pulse 8 GB graphics card. Based on a compacted variant of its RX 9060 XT Pulse board design, the card forms AMD's new entry-level in the RX 9000 series generation of GPUs powered by RDNA 4 graphics architecture. AMD carved the RX 9050 out of the 4 nm...
Leaks Point to Return of 4GB GPUs as AI Flash Memory Crisis Worsens
Welcome back to 2022.
Moonshot just gated new users from its new Kimi model
The Chinese AI startup said it saw massive demand over the last two days and had to pause new subscriptions to reduce strain on its GPUs.
Assassin's Creed Black Flag Resynced Performance Benchmark Review
Ubisoft's Assassin's Creed Black Flag Resynced breathes new life into the classic with modern graphics, gameplay improvements and full ray tracing support. In our performance review, we examine its graphical fidelity, VRAM usage, upscaling options, and how well it runs across a wide range of modern GPUs.
AMD FSR Multi-Frame Generation With Up to 8x Mode in Development
AMD has reportedly been testing a new and highly anticipated FSR multi-frame generation (MFG) technology for its Radeon RX GPUs. This information comes from a third-party software called RadeonTuner, which allows access to Radeon driver settings without using the AMD Software: Adrenalin Edition...
NVIDIA GPU Hotspot Temperature Has Been Unlocked Through Mods, & Shows Widespread Thermal Issues Affecting RTX 50 GPUs That Throttle Gaming Performance
Mods have unlocked NVIDIA's GPU Hotspot measurement, which can give you a clue why your latest card is running at lower performance.
NVIDIA announces 14 new GeForce cards, but none of them have a GPU
NVIDIA has announced 14 collectible GeForce Trading Cards based on classic GPUs, games and technology demos. They will be distributed through giveaways and gaming events.
Accidentally Leaked FSR 4.1.1 INT8 Reportedly Works With RDNA 3.5 Graphics
Valve's leak reveals FSR 4.1.1 INT8 supports RDNA 3.5 and RDNA 2 GPUs, with successful tests on RX 7800 XT and RX 6900 XT
Valve leaks FSR 4.1 for RDNA 3 and RDNA 3.5 GPUs through Proton Experimental
FSR 4.1 file works with RDNA 3 and RDNA 3.5 GPUs Valve appears to have pushed an early FSR 4.1 build through Proton Experimental.
AMD has reportedly released FSR 4 for RDNA3 GPUs in Valve Proton Experimental
AMD FSR4 appears to be coming to Valve’s Proton Experimental.
From wool sneakers to GPUs: Inside Allbirds' pivot to AI infrastructure
Allbirds announced It would no longer sell shoes and instead become an AI infrastructure provider. Business Insider talked to the new CEO.
South Korea to seek priority supply of Nvidia Vera Rubin GPUs, science minister says
South Korea will ask for priority supply of Nvidia's next-generation Vera Rubin graphics processing units, as deliveries are expected to be delayed, Science and ICT Minister Bae Kyung-hoon said on Monday.
Intel warns of "crazy" price: Arc G3 Extreme tramples AMD Ryzen Z2 Extreme for expensive price
Intel has detailed the performance and power efficiency of its upcoming Arc G3 Extreme SoC. Designed for handheld gaming consoles, the Intel Arc G3 and Arc G3 Extreme rely on Intel's latest Panther Lake CPU architecture and feature 10 and 12 Xe-core iGPUs, respectively. However, the consoles powered by these chips could be extremely expensive.
NVIDIA DLSS 4.5 Ray Reconstruction coming in August for RTX 20, 30, 40 and 50 series
NVIDIA confirms DLSS 4.5 Ray Reconstruction for all GeForce RTX GPUs NVIDIA has announced DLSS 4.5 Ray Reconstruction, a new version of its AI-based
Thermal Grizzly prepares WireView Pro II Noctua Edition for 12V-2×6 GPUs
Thermal Grizzly is turning WireView Pro II into a quieter Noctua version Thermal Grizzly is preparing a Noctua Edition of its WireView Pro II GPU power
007 First Light just got a second Nvidia promotion, as it now comes free with GeForce Now Ultimate
After launching its 007 First Light bundle for RTX 50 series GPUs, Nvidia is now giving away the game for free with GeForce Now.
Open-Source "low_latency_layer" Brings Reflex & Anti-Lag 2 To AMD & Intel GPUs On Linux
A new open-source project called low_latency_layer is an implicit Vulkan layer that enables AMD Anti-Lag 2 and NVIDIA Reflex 2 to reportedly work in a hardware-agnostic manner so that AMD and Intel graphics cards can both enjoy Reflex or Anti-Lag 2 working on non-AMD graphics cards as well.
AMD promises to bring improved, hardware-backed FSR 4 upscaling to older Radeon GPUs
FSR 4.1 running on RDNA3 or RDNA2 GPUs may take a bigger performance hit.
AMD FSR 4.1 Now Coming to Older Radeon 7000 GPUs in July After Radeon 9000 Debut, With Radeon 6000 Support Coming in 2027
The latest PSSR-influenced upscaler gets a wider release.
AMD FSR Upscaling 4.1 officially coming to Radeon RX 7000 GPUs in July, RX 6000 in 2027
Breaking, AMD confirms FSR 4.1 for Radeon RX 7000 series After months and often years of waiting, AMD is finally confirming support for FSR4 on RDNA3 and
Nvidia fixes the 8GB RAM problem with one of its GPUs—if you can pay for it
Framework charges nearly double for the 12GB version of the mobile RTX 5070.
Framework’s first eGPUs turn its laptop into a desktop PC
Only power users need apply — and you’ll have to shut down the laptop before you plug or unplug.
Valve Developer Further Improves Old AMD GPUs: HD 7870 XT Finally Working On Linux
Timur Kristóf of Valve's Linux graphics driver team is the one that worked on improving the old AMD Radeon GCN 1.0/1.1 graphics card support by making AMDGPU driver improvements so it could become the default for these Southern Islands and Sea Islands GPUs rather than the legacy Radeon kernel driver
Allbirds shares surge over 430% as footwear firm trades shoes for AI business
In a radical strategic overhaul, Allbirds is divesting its footwear business while securing $50 million (€42.4 million) in new financing to acquire GPUs and rebrand as NewBird AI.View on euronews
The largest orbital compute cluster is open for business
Kepler Communications is flying 40 GPUs in Earth orbit. And its latest customer is Sophia Space.
The largest orbital compute cluster is open for business
Kepler Communications is flying 40 GPUs in Earth orbit. And its latest customer is Sophia Space.
NVIDIA DLSS 5 — Everything We Know So Far About NVIDIA’s Latest Neural Rendering Technology
NVIDIA's DLSS 5 is set to launch Fall 2026, offering real-time neural rendering for photoreal graphics on RTX 50 Series GPUs.
4 Clever Uses For The Thunderbolt Port On Your PC
Many modern laptops and PCs now support Thunderbolt, which looks like a regular USB-C port but can support high-speed external drives, eGPUs, and more.
AMD or Nvidia eGPUs can work on Apple Silicon Macs, but not for graphic acceleration
Apple has signed a driver for AMD or Nvidia eGPUs connected to Apple Silicon but there are some big caveats, and it won't improve your graphics. Here's what they're for.
AMD Rolls Out Gemma 4 Model Support Across Full Range of GPUs & CPUs
AMD has rolled out official support for Google's Gemma 4 across its full range of GPUs & CPUs, offering support for the compact AI model.
New Rowhammer attacks give complete control of machines running Nvidia GPUs
Both GDDRHammer and GeForge hammer GPU memory in ways that compromise the CPU.
AMD Stock Slips as Gamers Call for FSR 4 Support on RDNA 3 and RDNA 2 GPUs
AMD stock fell on Friday as gamers call on the company to bring FSR 4 support to older GPUs.
Nearly one-third of comments under AMD’s 9950X3D2 launch video ask for FSR 4 on older GPUs
AMD, it's time to react This is probably not what AMD expected to see. AMD’s YouTube announcement for the Ryzen 9 9950X3D2 quickly turned into another
Intel Core Ultra 270K and 250K Plus review: Conditionally great CPUs
Great performance for the price, if you ignore the price of RAM, SSDs, and GPUs.
Intel launches Arc Pro B70 at $949 with 32GB GDDR6 memory
Intel launches Arc Pro B70 and B65 with 32GB GDDR6 featuring "Big Battlemage" GPU Intel has introduced two new Arc Pro workstation GPUs, the Arc Pro B70
Big Battlemage Is Here – Intel Unveils Arc Pro B70 & B65 GPUs, Up To 32 GB Memory & 367 TOPS For AI
Intel has finally unveiled its "Big Battlemage" GPUs, the Arc Pro B70 & Arc Pro B65, with up to 32 GB of memory for AI & Pro workloads.
Snapdragon 8 Elite Gen 6, 8 Elite Gen 6 Pro specifications leaked well before official announcement
Leak reveals Qualcomm Snapdragon 8 Elite Gen 6 chips with 2nm process, dual variants, LPDDR6 support, and upgraded GPUs. Read more about expected specifications here
The 3DFX Voodoo Lives Again In An FPGA
The 3DFX Voodoo was not the first dedicated 3D graphics chipset by any means, but it became the favourite for gamers among the early mass-market GPUs. It would be found on a 3D-processing-only PCI …
Pentagon’s $200B Iran war request rattles Congress
It comes as oil prices also rose above $119 per barrel
New dispute emerges around Subnautica 2’s early access release
Lawyers for the ousted—and now partially reinstated—leadership team of Unknown Worlds, the studio behind Subnautica 2, are crying foul over news that the game will be released into early access in May. The release announcement earlier this week was made improperly, potentially “damaging the game and sowing additional confusion among the Subnautica community,” the lawyers said in a letter sent on Tuesday to Vice Chancellor Lori Will of Delaware’s Court of Chancery. In their letter, the lawyers also floated the idea of asking Will to find Krafton in contempt of the court’s ruling. The letter was sent just one day after Will’s bombshell ruling that publisher Krafton had breached its contract with Unknown Worlds’ executive team, when it fired them last June. The new issue isn’t necessarily the May early access timeframe, but how the news was announced. The letter and Krafton’s response have not been previously reported. A quick recap of prior events: Subnautica 2 is an unreleased, underwater survival adventure and the most-wishlisted game on Steam. Last year, longtime Unknown Worlds CEO Ted Gill, along with studio co-founders Charlie Cleveland and Max McGuire, were fired by Krafton. They swiftly sued, saying the publisher had terminated them to avoid paying out up to $250 million in potential performance-based bonuses. The studio leaders had hoped to earn that with an early access release of the game, which Krafton had blocked. The publisher argued in court that Subnautica 2 wasn’t ready, which the ousted execs disputed. On Monday, Will ruled that Krafton had violated its contract, firing the execs to avoid the performance payouts, and that it had to reinstate Gill and give him full authority over any early access release for Subnautica 2. On Tuesday, IGN reported on a new memo issued that day to Unknown Worlds staff by Steve Papoutsis, the executive Krafton had put in charge of Unknown Worlds after the 2025 firings. In the note, Papoutsis said that Subnautica 2 had passed a key development milestone “last week” and that he and Krafton had determined the game “is ready for Early Access release in May.” This news appeared to catch the lawyers for Gill, Cleveland and McGuire by surprise. On Tuesday night, they sent a letter to Will expressing “serious concerns” about whether it would be possible to forge an agreement with Krafton consistent with the judge’s order. Papoutsis’ memo was a problem, they wrote. He didn’t have the authority to declare anything about Subnautica 2’s release. As of 9am ET on Monday, March 16, the effective date of Will’s order, Papoutsis wasn’t CEO of Unknown Worlds. Gill was. And, per Will’s order, only Gill could make Subnautica 2 release decisions. They continued: Krafton self-servingly announced the launch without any regard to its impact on the game, the team, or the community—let alone this Court’s Opinion. Announcing the release of a game is momentous, and it is typically accompanied by significant marketing activity, fanfare, and community coordination. And most importantly, the announcement is carefully designed to maximize excitement for the game. That entire process was supposed to be driven by Mr. Gill. However, in defiance of the Court’s Opinion, Krafton has now taken that away, further damaging the game and sowing additional confusion among the Subnautica community. The lawyers speculated that Krafton had “intentionally leaked” Papoutsis’ memo and then issued an official statement to confirm the leak. Game File is a reader-supported publication. My reporting is only possible through support from readers like you. To support my work, consider becoming a free or paid subscriber. Subscribe On Wednesday, lawyers for Krafton defended their client in their own letter to vice chancellor Will, saying nothing they did violated her order nor showed contempt: Papoutsis’s message was simply celebrating the UW employees’ efforts toward a past event: Krafton’s pre-Opinion determination that Subnautica 2 was ready for Early Access release. There was nothing improper about conveying the results of the milestone review or thanking the development team for their dedication and talent. They added that Gill could still make his own call about Subnautica 2’s release: In his role as CEO, Gill will be able to assess independently his views on the state of Subnautica 2 and the appropriate release schedule; nothing in Papoutsis’s message alters Gill’s authority or discretion. Vice chancellor Will has not responded to the letters and doesn’t necessarily have to. Both parties were previously ordered by Will to confer this week and submit a proposed judgment consistent with her ruling. The letter from Gill’s attorney’s this week doesn’t indicate whether he would or would not support a May release for Subnautica 2. But it could limit his options. And, if Gill’s lawyers are right, the lack of marketing fanfare around the leak of the May date could dull the launch of the game. Any harm to Subnautica 2’s launch is an extra sensitive issue given the context of the legal dispute: millions of dollars of payouts are on the line, tethered to the scale of the game’s initial performance. In her ruling on Monday, Will noted that there was obvious “bad blood” between Gill’s side and Krafton but that the two would need to work together: “They can—and must—act in good faith to navigate their remaining contractual relationship.” Item 2: A new kind of couch co-op game A highlight of GDC each year is the Alt.Ctrl showcase that features games played with extremely experimental controllers. Perusing this year’s offerings last Friday, I found a game controlled with giant scissors, another with big toothbrushes and one that amounted to dumpster diving. My favorite was Pivot, a co-op game about collectively maneuvering a couch. It was developed by students at the University of Utah’s masters of arts and engineering games program. Players lift and tilt the physical couch in order to control the position of a virtual one. That virtual one is moving through an obstacle course on-screen, while being pursued by a giant skull. Developer Nishant Verma told me at GDC that the couch is made of styrofoam (this became more evident as I watched!) and that the name was a riff on an episode of the sitcom Friends. Item 3: In brief… 🚫 Ubisoft is laying off more than 100 workers at its Red Storm studio, as the North Carolina location known for decades for its Tom Clancy games, ceases in-house game development, VGC reports. 👀 On Monday, Nvidia’s revealed its DLSS 5 AI-driven graphics technology that purports to improve game graphics… and it has been roasted about it by the media and the online gaming public for the remainder of the week. The tech changed the faces of characters in footage from Resident Evil Requiem, Hogwarts Legacy and EA Sports FC, drawing accusations of giving carefully art-directed games the generic look of AI “slop.” Nvidia CEO Jensen Huang had said on Monday that “DLSS 5 is the GPT moment for graphics—blending handcrafted rendering with generative AI to deliver a dramatic leap in visual realism while preserving the control artists need for creative expression.” By mid-week, Huang was telling Tom’s Hardware that the new tech’s critics were “completely wrong” because game-makers could still “fine-tune the generative AI” to suit their game’s aesthetics.
Intel Precompiled Shaders Accelerate Game Load Times By Up To 3x On Arc GPUs, Now Available & Supported In 13 Games
Intel Arc GPUs can finally enable Precompiled Shaders for faster loading times that are supported in 13 games at launch.
Intel enables precompiled shader selivery on Arc B-series and Core Ultra iGPUs in new driver, up to 3x faster loading times
Intel Arc gets Graphics Shader Distribution Service, 13 games supported at launch Today, Intel has rolled out its Graphics Shader Distribution Service in
Scientists used 7,000 GPUs to simulate a tiny quantum chip in extreme detail
Researchers have pushed quantum chip design into a new era by simulating every physical detail before fabrication. Using a supercomputer with nearly 7,000 GPUs, they modeled how signals travel and interact inside an ultra-tiny chip. Unlike earlier “black box” approaches, this method captures real materials, layouts, and qubit behavior. The result is a powerful new way to spot problems early and build better quantum hardware faster.
Open-Source "GreenBoost" Driver Aims To Augment NVIDIA GPUs vRAM With System RAM & NVMe To Handle Larger LLMs
An open-source, independently developed Linux kernel module called GreenBoost aims to augment the dedicated video memory on NVIDIA discrete GPUs with system memory and NVMe storage
NVIDIA’s Answer to the GPU Shortage Has Samsung Resurrecting the GeForce RTX 3060, and It Might Just Work
NVIDIA now plans to bring older GPUs back to life, and Samsung Foundry is preparing to restart production lines for the GeForce RTX 3060.
NV-UV brings one-click undervolting to GeForce RTX 50 GPUs
NV-UV is a new GeForce RTX 50 undervolting tool built on MSI Afterburner A new utility called NV-UV has surfaced for NVIDIA GeForce RTX 50 series graphics
Yet Another Fix Coming For Older AMD GPUs On Linux - Thanks To Valve Developer
Timur Kristóf of Valve's open-source Linux graphics driver team has been doing a fantastic job enhancing the older AMD Radeon GPU support under Linux
Intel releases XeSS 3 with Multi-Frame Generation, now rolling out to Arc GPUs
Intel’s new Arc driver enables XeSS 3 Multi-Frame Generation Intel has started shipping XeSS 3 support with Multi-Frame Generation in its latest Windows
Nvidia Stock (NVDA) Rallies on Reports of Increased GPU Prices
Nvidia stock gained on reports that the company has increased the price of its GPUs.
Screens on GPUs simply aren't enough, ZALMAN puts a screen on VGA support
ZALMAN puts an LCD screen on a GPU support, costs only $12 PC cases and coolers keep adding small displays. ZALMAN is now putting one on a GPU anti-sag
ASUS changes mind, will continue selling the RTX 5070 Ti after all
After reportedly telling Hardware Unboxed that it was putting its RTX 5060 Ti 16GB and 5070 Ti GPUs into "end-of-life status," ASUS is backtracking.
NVIDIA RTX 50 GPU Shortages: Green Team Realigns Production of Entire Gaming Stack, While ASUS Says RTX 5070 Ti Production Has Halted
It appears that NVIDIA is finally killing some of the GeForce RTX 50 series cards such as the RTX 5070 Ti and may cut supply for more GPUs.
NVIDIA rolls out DLSS 4.5 to all RTX GPUs
The company announced the latest version of its image-upscaling tech at CES last week.
NVIDIA Prioritizes RTX 5060 Series GPUs Over Rest of Lineup, RTX 5090s Selling Above $3500
NVIDIA has put the majority of its RTX 50 lineup on the back foot as it now prioritizes the RTX 5060 series over everything else.