RSS Reader

Intel's newest contribution to the upstream LLVM compiler stack is XeVM as their Multi-Level Intermediate Representation "MLIR" dialect catering to modern Intel graphics processors.

The MLIR intermediate representation is important for more optimizations, being able to represent deep learning graphs by the likes of TensorFlow, and more with its ability to support domain-specific compilers and better handle heterogeneous compute hardware compared to alternative IRs. MLIR is becoming very important within the industry and is also being worked on by AMD as part of their unified AI software stack

Kernel Stack Watch Proposed As New Linux Debugging Tool

Phoronix | Aug. 19, 2025, 5:55 a.m. | Original

Patches were posted on Monday for Kernel Stack Watch, a new lightweight debugging tool for detecting kernel stack corruption in real-time on Linux.

Jinchao Wang announced the creation of Kernel Stack Watch (kstackwatch) as this new debugging tool for uncovering kernel stack corruption. He summed up nicely with yesterday's patches:

"This patch series introduces Kernel Stack Watch (KSW), a lightweight debugging tool for detecting kernel stack corruption in real-time. The motivation comes from cases where corruption happens silently in one function but only

New Linux Patches Allow Manipulating Out-Of-Memory Behavior Using BPF

Phoronix | Aug. 18, 2025, 8:47 p.m. | Original

Google engineer Roman Gushchin has proposed the ability for the Linux kernel to customize the out-of-memory "OOM" behavior using BPF programs.

While there is the likes of systemd-oomd for a user-space out-of-memory killer service, Linux memory management expert Roman Gushchin of Google has proposed the ability for (e)BPF programs to customize the OOM behavior within the Linux kernel.

Under the patches proposed today, the Linux kernel out-of-memory handling policy could be manipulated via BPF programs

AMD "GFX1250" To Double The Number Of User SGPRs

Phoronix | Aug. 18, 2025, 7:34 p.m. | Original

We still don't have much confirmed information on AMD's GFX1250 that has come to light in recent months due to being developed for their AMDGPU shader compiler back-end within LLVM. AMD GFX1250 surfaced in LLVM activity and the past few months has been seeing more additions made. A notable change committed today for LLVM is that the number of user Scalar General-Purpose Registers (SGPRs) has doubled.

AMD GFX1250 from the code is noted as being

AMD ZenDNN 5.1 Released For Enhancing AI Inference Performance On EPYC CPUs

Phoronix | Aug. 18, 2025, 4:18 p.m. | Original

Following the AMD ZenDNN 5.0 release from last year's EPYC Turin launch that brought big performance improvements for CPU-based inferencing with this open-source library compatible with Intel's oneDNN, today marks the availability of ZenDNN 5.1 as the next update.

ZenDNN 5.1 is now available for this AMD Zen optimized Deep Neural Network Library for CPU-based AI inferencing. ZenDNN 5.1 is optimized for use with TensorFlow 2.19 and PyTorch 2.7.

With ZenDNN 5.

Kdenlive 25.08 Preps For Future Hardware Acceleration Features

Phoronix | Aug. 18, 2025, 3:24 p.m. | Original

Kdenlive 25.08 is out today as the newest feature release for this KDE/Qt-aligned open-source video editor. In addition to a number of new features, there are also many bug fixes including more than 15 crash fixes.

Kdenlive 25.08 overhauls its audio mixer interface to provide better visuals and also fixing issues for those making use of HiDPI displays.

The Kdenlive titler has also received improved support for SVGs and images, improved Scope styling, various sub-title / text

Git 2.51 Released With More Changes Preparing For Git 3.0

Phoronix | Aug. 18, 2025, 1:17 p.m. | Original

Git 2.51 is out this Monday as the newest milestone for this widely-used, distributed version control system.

As covered in the 2.51-rc0 announcement earlier this month, Git 2.51 comes with more preparations for Git 3.0 . Git 3.0 is the point at which the default hash function will change from SHA-1 to SHA-256 for greater repository security against potential hash collisions.

Git 2.51 also delivers on cruft-free

A Deep Dive Into The Power & Thermals For The Framework Desktop With AMD Ryzen AI Max

Phoronix | Aug. 18, 2025, 11:40 a.m. | Original

The Framework Desktop is a nifty and powerful mini PC powered by AMD Ryzen AI Max "Strix Halo". It's been a pleasure testing this small yet powerful Linux-friendly system that easily offers much better performance than the Intel Core Ultra 9 and superb energy efficiency . For complementing the data shared earlier this month in our Framework Desktop review , today's article is a deep dive into the power and thermals of the Framework Desktop in a few different configurations.

AMD Posts Latest Linux Patches For Improving S5 Power Consumption

Phoronix | Aug. 18, 2025, 11:02 a.m. | Original

One of the areas worked on by AMD Linux engineers recently to better optimize the Linux kernel on AMD Ryzen platforms is for lowering the power consumption in S5 state due to some devices not being put into a low power state when the system is powered off. Sent out today was the sixth iteration of these patches.

AMD engineer Mario Limonciello on Sunday posted the sixth iteration of these Linux kernel patches to help with the S5 power consumption. Mario explained with the patch cover letter:

"A variety of issues both in

Open Source LLM

Firefox 142 Now Available - Allows Browser Extensions/Add-Ons To Use AI LLMs

Phoronix | Aug. 18, 2025, 9:36 a.m. | Original

The Firefox 142.0 release binaries are now available ahead of the official release announcement due out on Tuesday. Firefox 142 isn't bringing many notable changes but one is likely to cause some contention around Firefox Extensions.

Firefox 142 isn't particularly exciting for this August web browser update but does improve the drag-and-drop support for blob images, better scrolling speed in the bookmarks dialog, and some other refinements. For web developers Firefox 142 brings the Prioritized Task Scheduling

Intel Upstreams XeVM Into LLVM

Kernel Stack Watch Proposed As New Linux Debugging Tool

New Linux Patches Allow Manipulating Out-Of-Memory Behavior Using BPF

AMD "GFX1250" To Double The Number Of User SGPRs

AMD ZenDNN 5.1 Released For Enhancing AI Inference Performance On EPYC CPUs

Kdenlive 25.08 Preps For Future Hardware Acceleration Features

Git 2.51 Released With More Changes Preparing For Git 3.0

A Deep Dive Into The Power & Thermals For The Framework Desktop With AMD Ryzen AI Max

AMD Posts Latest Linux Patches For Improving S5 Power Consumption

Firefox 142 Now Available - Allows Browser Extensions/Add-Ons To Use AI LLMs

Menu