SCROLL NEWS / DISCOVERY
Search headlines
Find the stories shaping the conversation.
Results for " kernel " 82 found 🔀 AI-powered Shuffle
Sign in to save shuffle history.
Chinese state hackers use rootkit to hide ToneShell malware activity
A new sample of the ToneShell backdoor, typically seen in Chinese cyberespionage campaigns, has been delivered through a kernel-mode loader in attacks against government organizations.
Linus Torvalds: Someone ‘more competent who isn't afraid of numbers past the teens’ will take over Linux one day
: Emperor Penguin releases kernel 7.0 rc1 with some numerological musings
Linux 7.1-rc4 Released With Many Fixes, New Documentation For Security/AI Topics
It was another busy week in the Linux 7.1 kernel space that has culminated with the release of Linux 7.1-rc4.
Linus Torvalds says AI has made 'huge' Linux kernel updates the new normal
He’s not thrilled, but won’t let it delay the release of version 7.2
Exploit available for new DirtyDecrypt Linux root escalation flaw
A recently patched local privilege escalation vulnerability in the Linux kernel's rxgk module now has a proof-of-concept exploit that allows attackers to gain root access on some Linux systems.
Linux To Remove ISA Speech Synthesizer Driver That Likely Hasn't Been Used In Decades
Following the process of phasing out Intel 486 CPU support and other old hardware drivers that were dropped in the Linux 7.1 kernel cycle for reducing the kernel maintenance burden, the upcoming Linux 7.2 cycle is continuing the trend of phasing out some of the old hardware support that is very obsolete, likely having no users on the latest upstream kernels, and no one formally maintaining the obsolete drivers.
KernelScript: A Programming Language For Kernel Customization & App Optimizations
Multikernel Technologies Inc has been working on a multi-kernel architecture for the Linux kernel while in addition to that they have been developing KernelScript as a domain-specific language for carrying out Linux kernel customizations and app-specific optimizations.
Next iPad Mini With OLED Display Now Rumored to Feature A20 Pro Chip
Apple's next-generation iPad mini will be equipped with an A20 Pro chip, according to a MacRumors tipster who analyzed a macOS kernel debug kit containing internal Apple codenames. The kit was accidentally released on Apple's website earlier this year, but it was quickly pulled after information started leaking out of it.
Cinemark theaters to offer $5 Bring Your Own Bucket popcorn deal
Calling all kernel connoisseurs: Cinemark is reprising its Bring Your Own Bucket popcorn deal, and you'd \
Linux Kernel 7.2 Officially Released, This Is What’s New
Linux kernel 7.2 is now available to download with new features, improvements, and enhanced hardware support through new and updated drivers.
Linux 6.19-rc1 Released From Japan
The Linux 6.19-rc1 kernel is out to cap off the Linux 6.19 merge window
Linux 7.2 Released With Faster I/O, New AMD & Intel Driver Improvements
Linus Torvalds just released Linux 7.2 as stable! After a very busy kernel cycle due to the increased patch and reporting churn due to AI/LLM ages, Linux 7.2 managed to make it out today on-schedule for this feature-packed kernel
RADV Driver Lands Another Big Improvement For Early AMD GCN Graphics Cards
Beyond Linux 6.19 switching old AMD GCN 1.0 and 1.1 GPUs to the AMDGPU kernel driver by default for better performance, RADV out-of-the-box, and more, there are still more improvements planned for these aging AMD graphics cards
Linux 7.3 Better Protects Against Buggy EFI Firmware Taking Down The Rest Of User-Space
The EFI updates were merged today for the in-development Linux 7.3 kernel
The third major Linux kernel flaw in two weeks has been found
AI is exposing Linux security holes faster than developers can patch them. Fragnesia is the latest. Here's what we know about it.
A Lot Of Exciting Changes To Look Forward To With Linux 6.20 -- Or Linux 7.0
With Linux 6.19 due for release later today it then opens up the next kernel merge window
Cleveland’s Independent Music and Comedy Venues Drive $742.5 Million in Economic Impact, New Study Finds
A kernel from the study: The city’s venues account for nearly 50 percent of the indie live entertainment GDP of the Buckeye State, more than 28 states.
Linux 7.0.8 Released & LTS Kernels Updated For ssh-keysign-pwn
Following yesterday's disclosure of the ssh-keysign-pwn vulnerability that allows unprivileged users to read root-owned files, a slew of new stable kernel releases are out today to address this latest Linux security issue.
The New Linux Kernel AI Bot Uncovering Bugs Is A Local LLM On Framework Desktop + AMD Ryzen AI Max
Earlier this month on Phoronix we were the first to draw attention to a new fuzzing tool / AI bot uncovering kernel bugs by Greg Kroah-Hartman, the 'second in command' for Linux kernel development and stable maintainer
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.
Intel Xe vs. i915 Driver Performance On Linux 6.19 For Arc Alchemist GPUs
Similar to AMD GCN 1.0/1.1 GPUs where there was product overlap between the Radeon and AMDGPU kernel drivers (and now using AMDGPU by default for those aging Radeon GPUs with Linux 6.19), the Intel Arc A-Series 'Alchenist' graphics cards are in a similar boat.
AWS Engineer Reports PostgreSQL Performance Halved By Linux 7.0, But A Fix May Not Be Easy
An Amazon/AWS engineer raised the alarms on Friday over the current Linux 7.0 development kernel leading to the throughput for the PostgreSQL database server being around half that of prior kernel versions
AMD RDNA 4m Firmware Published For Linux Ahead Of Launch
In addition to AMD's Friday afternoon big code drop of all the UALink enablement patches for the Linux kernel and the AMDGPU driver, there was another big drop from AMD to end out the week..
Update Now: iOS 26.6 and macOS Tahoe 26.6 Patch Hundreds of Security Flaws
Apple today released iOS 26.6, iPadOS 26.6, and macOS Tahoe 26.6, all of which have a long list of security fixes. iOS 26.6 and iPadOS 26.6 address almost 90 security vulnerabilities affecting everything from the App Store to the Neural Engine. Multiple kernel and WebKit vulnerabilities were fixed, along with problems affecting Wi-Fi, Siri, and the iPhone's image processing. Details on all of the fixes can be found in Apple's security support document.
Early Linux 6.19 Benchmarks On AMD EPYC 9965 2P Excelling For AI & HPC Performance
As the Linux 6.19 merge window winded down this weekend, I began running this development kernel on more systems.
The Linux kernel now allows AI-written code, but you're on the hook for it
Your AI-generated code is still your code.
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
Is it true that … you need to take 10,000 steps a day to be healthy?
The fitness benchmark has surprisingly unscientific origins – yet it does contain more than a kernel of truth
Linus Torvalds Working On A Hefty Linux Kernel Improvement With Claude Code
Linux creator Linus Torvalds has been known to use AI to help with kernel debugging and outside of the kernel tree has also done some vibe coding with Google's Antigravity
New Fragnesia Linux Kernel LPE Grants Root Access via Page Cache Corruption
Fragnesia CVE-2026-46300 corrupts Linux page cache via XFRM ESP-in-TCP, enabling local root access on major distros.
NVIDIA Blackwell-Next name spotted in Linux kernel patch
NVIDIA engineers are already preparing Linux support for hardware listed as “Blackwell-Next.” The name appears in a Linux kernel patch for the NVGrace-GPU VFIO…
Cinemark theaters to offer $5 Bring Your Own Bucket popcorn deal
Calling all kernel connoisseurs: Cinemark is reprising its Bring Your Own Bucket popcorn deal, and you'd "butter" believe it's "a-MAIZE-ing."
Two More Studio Display 2 Upgrades Leaked in New Report
Macworld's Filipe Espósito today again reported that Apple appears to be developing two new Studio Display models with different specs. Based on lines of code in "internal Apple files," which likely refers to a macOS Tahoe Kernel Debug Kit that leaked online last year, Espósito continues to believe that both of the new Studio Display models will feature ProMotion, enabling up to a 120Hz refresh rate, as well as HDR support for increased brightness and dynamic range.
Linux 7.2 Surpasses More Than 43 Million Lines In The Kernel Tree
Today marks the last day of the Linux 7.2 merge window with Linux 7.2-rc1 due out later today
9-Year-Old Linux Kernel Flaw Enables Root Command Execution on Major Distros
CVE-2026-46333 is a nine-year Linux kernel improper privilege management flaw introduced in November 2016 with a CVSS score of 5.5.
Fact check: Trump keeps making stuff up about Canada
President Donald Trump’s criticism of the US neighbor has been filled with lies – a remarkable variety of assertions that either wildly distort a kernel of truth or are just made up.
Linux Addressing Out-Of-Memory Killer Inaccuracy On Large Core Count Systems
A patch is on the way to the Linux kernel and looks like it could be ready for the 6.20~7.0 kernel for addressing out-of-memory 'OOM' killer inaccuracy behavior when dealing with large core count systems.
Linux Kernel Dirty Frag LPE Exploit Enables Root Access Across Major Distributions
Dirty Frag exposes Linux systems to root escalation through chained kernel flaws, impacting Ubuntu, RHEL, Fedora, and others.
New Fragnesia Linux flaw lets attackers gain root privileges
Linux distros are rolling out patches for a new high-severity kernel privilege escalation vulnerability (known as Fragnasia and tracked as CVE-2026-46300) that allows attackers to run malicious code as root.
Linux kernel maintainers are following through on removing Intel 486 support
Linux devs think even one second spent on 486 support is a second too many.
Linux 6.19 Features Include Many Benefits For Intel & AMD Users
With the Linux 6.19 stable kernel expected to be released tomorrow (8 February), here is a reminder about the top features to expect from this next version of the Linux kernel.
Google's Gemini Partially Figures Out A Lengthy Linux Boot Time On Modern ASUS Laptop
Google Antigravity with the Gemini 3.5 Flash model helped a Linux user sort out a situation where his laptop was taking around 36 seconds to boot the kernel, which shouldn't be the case for the high-end laptop with AMD Ryzen 9 processor and 32GB of RAM
Linux 7.0-rc5 Released: Linux 7.0 "Starting To Calm Down"
Linus Torvalds just issued Linux 7.0-rc5 as we inch toward the stable Linux 7.0 kernel release in April.
Linux 6.19 Lands Fix For Dead WiFi With MediaTek MT792x Wireless
Merged to Linux Git on New Year's Eve was a fix in the form of a code revert for broken MediaTek WiFi on the in-development Linux 6.19 kernel.
Linux's Latest Vulnerability Allows Reading Root-Owned Files By Unprivileged Users
Following Dirty Frag, Fragnesia, and other Linux kernel vulnerabilities making themselves known in recent days, the latest now is ssh-keysign-pwn.
Linux Finally Eliminates The strncpy API After Six Years Of Work, 360+ Patches
Linux 7.2 has finally eliminated the strncpy API from the Linux kernel
Leaker Says Apple's Lower-Cost MacBook Will Have These 8 Limitations
Apple is expected to unveil its long-rumored lower-cost MacBook next week. Given it will be more affordable, this MacBook model will obviously have some reduced specs and compromises compared to the MacBook Air and MacBook Pro. While we are still waiting for Apple to announce the new MacBook, a leaker has shared eight alleged limitations to expect, based on an internal version of Apple's Kernel Debug Kit for a macOS Tahoe beta that leaked online last year.
NVIDIA Graphics On Haiku OS Make Progress With NVIDIA Open Kernel Modules + NVK/Zink
As a wonderful New Year surprise, there's good momentum on NVIDIA graphics support for the BeOS-inspired Haiku open-source operating system.
Apple Tested a MacBook With the A15 Chip
An internal Apple kernel debug kit suggests Apple has tested a MacBook with the A15 chip, alongside a separate A18 Pro-based MacBook that appears to be closer to a shippable product. The information comes from internal kernel debug kit files used by Apple engineers. The kit was accidentally released on Apple's website earlier this year, but it was quickly pulled after information started leaking out of it.
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
First Public macOS Kernel Exploit on Apple M5 Prepared Using Mythos Preview in Five Days
Apple's M5 silicon has reportedly been exploited for the first time in a public macOS kernel memory corruption attack, successfully bypassing the company's notable hardware-level memory protection.
Valve's Linux Efforts, Kernel Improvements & KDE Plasma Wayland Advancements Topped 2025
After looking yesterday at the most viewed Linux hardware reviews and benchmarks of 2025, today's look is at the most popular open-source/Linux news of the past year
Several Linux Kernel Vulnerabilities Accidentally Posted To Public Kernel Mailing List
Seemingly as an accident, overnight several new Linux kernel vulnerability reports were posted to the public Linux kernel mailing list rather than the private security mailing list
Linux Finally Catches Up to Windows with a Game-Changing Performance Feature
For years, the Linux kernel’s scheduler has been world‑class at balancing loads, yet it missed a crucial, cache‑aware instinct. In modern multi‑core systems, that gap can turn into measurable latency, especially when threads bounce between cores that don’t share the same cache. A new upstream feature, often called Cache Aware Scheduling, is now set to change that equation. A cache‑savvy scheduler arrives At the heart of any OS, the scheduler decides which thread runs where and for how long. In modern CPUs, small private caches (L1 and L2) sit beside each core, while a larger Last Level Cache (LLC, typically L3) is shared among groups of cores. When a task migrates across LLC boundaries, its warm data may vanish from the cache, forcing slower trips to main memory. Cache Aware Scheduling keeps related tasks close to their shared LLC, reducing destructive migrations. By respecting cache topology, the kernel minimizes cold‑start penalties and preserves locality that many workloads desperately need. The result is less thrashing, fewer memory stalls, and more consistent throughput. "Keep tasks close to their data, and the system will keep performance close to its peak." Why this narrows the Windows gap Windows has long leaned on topology‑aware, cache‑sensitive heuristics, especially since the Windows 10 era. That advantage helped Microsoft handle hybrid designs with P‑cores and E‑cores, plus complex cluster and NUMA layouts. With cache awareness integrated upstream, Linux brings parity to this vital dimension, without sacrificing its trademark flexibility. The Linux approach remains deeply configurable, reflecting the ecosystem’s breadth across servers, desktops, and embedded devices. It layers atop existing NUMA‑balancing and energy‑aware logic, refining placement rather than reinventing the wheel. Crucially, it aligns scheduling with the hardware’s real shape, not just with abstract CPU counts. Real‑world gains and who benefits Early tests on Intel Sapphire Rapids platforms point to gains in the 30–45% range for select workloads. Those wins appear in cache‑sensitive tasks like in‑memory analytics, high‑thread compilation, and microservices with tight working‑sets. Games and latency‑bound engines can also feel steadier frame‑times, especially when threads share hot assets. The benefits extend to AMD’s 3D V‑Cache parts, where keeping threads near enlarged cache slices prevents needless misses. Hybrid x86 designs with performance and efficiency cores further profit when cache and core roles are scheduled in concert. Even handhelds running SteamOS can squeeze more out of limited power, translating cache locality into smoother play. Key advantages include: Lower end‑to‑end latency Fewer expensive misses to RAM Better multi‑socket and NUMA behavior Improved energy efficiency More predictable QoS under mixed loads Stronger scaling on dense servers Caveats, tuning, and rollout No scheduler change is purely free, and trade‑offs still exist. Favoring locality can reduce cross‑cluster balance if limits are set too tight. Likewise, fairness and utilization must remain healthy, especially under heterogeneous, bursty workloads. Linux’s implementation is designed to be measured and tunable, not a blunt instrument. It cooperates with energy‑aware scheduling for mobile efficiency, and with NUMA‑balancers on big iron. Expect architecture‑specific refinements as vendors surface richer topology hints and cache‑sharing maps. Upstream integration has begun, with broad distribution adoption likely over the next cycles. Many users will encounter the change through standard kernel updates in 2025–2026, depending on distro cadence. Server‑class kernels may enable or tune it sooner for targeted fleets where cache locality is money in the bank. How to think about performance impact Cache Aware Scheduling amplifies gains when your workload is both CPU‑bound and cache‑sensitive. Think tight inner loops, hot code paths, and datasets that fit snugly in shared LLC. It’s less dramatic for I/O‑bound services or memory‑hungry tasks that blow past cache capacity. Still, even modest locality improvements often translate into smoother tail‑latency, which is where user experience and SLAs typically break. Developers can further help by pinning related threads, batching work, and aligning data to reduce cross‑cluster chatter. The scheduler’s new awareness works best when applications expose good hints. The bottom line By understanding and honoring cache topology, Linux removes a subtle yet costly bottleneck. The kernel’s smarter placement keeps hot data hot and work near where it belongs. That narrows a long‑standing gap with Windows, while preserving the openness and tunability Linux champions. For gamers, creators, and operators, the payoff is practical: more consistent frames, faster builds, and snappier services. For the ecosystem, it’s another step toward hardware‑savvy scheduling that turns transistor complexity into real‑world speed. And for Linux itself, it’s a timely upgrade that matches today’s silicon with tomorrow’s expectations.
Snadragon X Elite Laptop Performance On Linux Ends 2025 Disappointing
As part of my various end-of-year benchmarking comparison articles for looking at the performance evolution of Linux is a fresh look at the Qualcomm Snapdragon X Elite laptop experience when using Ubuntu 25.10 with the latest X1E Concept packages, which includes taking the X1 Elite optimized kernel to the latest Linux 6.18 stable series.
Linus Torvalds "Did What I Preach", Used AI To Make Code Decipherable & Uncovered More Bugs
After using Claude Code to draft code to help make a fundamental improvement to the Linux kernel, Linux creator Linus Torvalds continued using AI this weekend and talking up its benefits
Linux 6.19-rc5 Brings Fix For Newer NVIDIA GPUs, Logitech HID++ For Anywhere 3S & Fixes
In addition to Linus Torvalds doing some vibe coding and more with his new 'AudioNoise' project this week, Linux 6.19 kernel development ticked back up with the holidays having passed
iPhone 17e and iPhone 18 Pro Modem Specs Leak Online
Internal Apple code suggests the iPhone 17e may ship with Apple's first-generation C1-series cellular modem technology while omitting Apple's N1 wireless chip, and it indicates Apple is still deciding between the C1X and C2 for the iPhone 18 Pro. The details come from internal kernel debug kit files derived from an Apple prototype device running an early build of what became iOS 26. The considerable leak has revealed specific details about dozens of upcoming Apple devices.
Linux 6.19 Kernel Benchmarks With X86_NATIVE_CPU Optimization
Added to the Linux kernel earlier this year was the new X86_NATIVE_CPU Kconfig option to enable compiler optimizations for the local/native CPU in use when building the Linux kernel
The 4th Linux kernel flaw this month can lead to stolen SSH host keys
The good news is there's already a patch. The bad news is that the fix isn't available for all Linux distributions yet. Here's what you can do in the meantime.
NVIDIA's Quest For A "Safe" Linux Kernel For Automobiles, Robotics
NVIDIA engineer Igor Stoppa presented at the Linux Plumbers Conference (LPC) earlier this month around using Linux in safety-critical environments like automobiles and the current shortcomings of the upstream Linux kernel and the challenges on achieving Automotive Safety Integrity Level (ASIL) certifications around the Linux kernel
‘There can be kernels of truth in crude stereotypes’
Opinion, comment and editorials of the day
Two CVSS 9.0 Linux KSMBD RCE Flaws Disclosed on Public Mailing Lists Before Kernel Fixes
Two CVSS 9.0 flaws in Linux KSMBD could enable remote code execution; one needs no authentication, and neither has a mainline fix.
Linux 7.0 Released With New Hardware Support, Optimizations & Self-Healing XFS
As expected the stable Linux 7.0 kernel was just released today in marking this next kernel release
Linus Torvalds keeps his ‘fingers and toes’ rule by decreeing next Linux will be version 7.0
: But first, kernel 6.19 is upon us, with many goodies
Linux 7.1 Released: New NTFS Driver, Intel FRED For Panther Lake, Faster Arc Graphics
Linus Torvalds just released the stable Linux 7.1 kernel and it's coming a half-day early thanks to his travel plans
iOS 27 and iOS 26.7 Fix More Than 100 Security Bugs
Today's iOS 27 update addresses over 100 security vulnerabilities, which is a good reason to install the new software. If you don't want to update to iOS 27 yet, Apple also released iOS 26.7 with over 80 security fixes. 75 of the fixes are shared across both updates, so iOS 26.7 addresses many of the same issues. There are multiple kernel fixes, including one that could allow a malicious app to gain root access and another that could allow for Bluetooth remote code execution.
Linux 7.1 Expected To Begin Removing i486 CPU Support
It's finally time: a patch queued into one of the development branches ahead of the upcoming Linux 7.1 merge window is set to finally begin the process of phasing out and ultimately removing Intel 486 CPU support from the Linux kernel
Kernel Graphics Driver Changes Already Begin Lining Up For Linux 6.20~7.0
Even before the Linux 6.19 merge window wrapped up this weekend with the Linux 6.19-rc1 release, there was already the first pull request to DRM-Next of the first batch of new material to be queued for Linux 6.19's successor.
Linus Torvalds Confirms The Next Kernel Is Linux 7.0
Following Linus Torvalds releasing Linux 6.19 stable, Linus Torvalds is now out with his customary release announcement
Sauer Brands offloads spice business to Watkins
Watkins will secure the branding and distribution rights for spices and seasonings under the Kernel Season’s, Spice Hunter and Sauer’s brands.
Linux Kernel 7.1 Officially Released, Here's What’s New
Linux kernel 7.1 is now available for download with new features, enhanced hardware support through new and updated drivers, and more.
New Runtime Standby ABI Proposed For Linux Akin To Microsoft Windows' "Modern Standby"
An exciting post-Christmas patch series out on the Linux kernel mailing list this morning is proposing a new runtime standby ABI that is similar in nature to the 'Modern Standby' functionality found with Microsoft Windows.
Linux Kernel Highlights For 2025: Schedulers, Rust & Torvalds' Commentary
With the end of the year quickly drawing to a close, here is a look back at the most-viewed Linux kernel news of 2025.
Linux 7.2-rc6 Released: "This RC Is Huge"
Like clockwork, the Linux 7.2-rc6 kernel is now available for testing in working toward the stable Linux 7.2 kernel in hopefully two week's time.
Fedora Continued At The Forefront Of Upstream Linux Innovations In 2025
Fedora Linux this year continued in punctually shipping the very latest upstream Linux innovations from the freshest Wayland components to Linux kernel features and continuing to leverage other improvements in the open-source world.
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
Linux 7.1 Removes Drivers For Long Obsolete Input Hardware: Bye Bus Mouse Support
Beyond Linux looking to remove old drivers due to the surge of AI/LLM bug reports, the Linux 7.1 kernel is also removing some old hardware drivers simply on the basis of long obsolete hardware
Linux 7.3 To Introduce DRM "Color Format" Property With AMD GPU Driver Support
While the Linux 7.2 kernel merge window is only ending later today to cap off the feature work on this next version of the Linux kernel, already for the Linux 7.3 kernel cycle later in the year there is one notable feature on the way: the DRM color format property is being introduced and being first supported by the AMDGPU kernel graphics driver.
AMD Linux Graphics Driver Introducing "Power Module" To Better Match Windows Behavior
With the set of today's AMDGPU kernel graphics driver Display Core (DC) patches is a rather curious addition with wiring up the Linux code to a 'power module' that looks like it will better match Microsoft Windows behavior with the AMD Radeon driver around display-related power savings features.
Apple Developing iMac Pro With M5 Max Chip
Apple is developing a high-end iMac featuring the M5 Max chip, according to information from leaked internal software. The finding comes from leaked kernel debug kit files used by Apple engineers. These kernel debug kit files enumerate unreleased Apple hardware by internal identifiers, such as codenames and platform names, and they can also include separate references that map platform identifiers to chip codenames and marketing names.
Fragnesia Made Public As Latest Linux Local Privilege Escalation Vulnerability
Following last week's disclosure of the Dirty Frag vulnerability for the Linux kernel, which only finished being patched up in mainline on Monday, Fragnesia is now public as a similar local privilege escalation (LPE) vulnerability.