Software Linux Reviews & Articles

There have been 981 Linux hardware reviews and benchmark articles on Phoronix for software. Separately, check out our news section for related product news.

The long-in-development work on Cache Aware Scheduling looks like it will come to a head soon with it looking like Cache Aware Scheduling will land for Linux 7.2. Ahead of the upcoming merge window I ran some fresh benchmarks looking at different areas where this feature is shining.

The Linux 7.1 kernel performance has been looking quite good on the various Intel/AMD systems I have tested over the past three weeks. Linux 7.1 does bring some solid improvements over Linux 7.0 prior in different workloads and haven't encountered any worrisome regressions compared to the current Linux 7.0 stable kernel. For those wondering the longer-term picture, here are benchmarks of Linux 7.1 Git compared to recent Linux LTS kernel series going back to 2023 for providing a picture at how the upstream Linux kernel has netted 13% faster performance (geo mean) on the same hardware in less than three years.

GCC 16.1 released at the end of April as the latest major, annual feature release to the GNU Compiler Collection. Early benchmarks showed some nice leads for GCC 16 over GCC 15. Continued testing of the new GCC 16 compiler has continued to show overall better performance of the resulting binaries than using GCC 15 on the same hardware and same compiler flags. That led many to wonder about the GCC 16 performance up against the latest LLVM/Clang open-source compiler, which is the focus of today's benchmarking showdown.

The Linux 7.1 development kernel that amounts to nearly 40 million lines has a lot of new features and changes in tow. While Linux 7.1 stable won't be out until mid-June, here is a look at the interesting changes coming with this next stable version of the Linux kernel.

With the GCC 16.1 compiler released last Thursday, I have begun running more compiler benchmarks on this first GCC 16 stable feature release. GCC 16 comes heavy on new changes in being the annual feature release and delivering changes from AMD Zen 6 and Arm AGI CPU support to new C++ features and even the Algol 68 programming language front-end. It's also looking quite good in the performance department relative to the GCC 15 compiler from last year.

With the Linux 7.1 merge window winding down ahead of the planned Linux 7.1-rc1 release on Sunday, I have begun testing out the Linux 7.1 Git state on various systems in my lab. So far Linux 7.1 appears to be looking good in the performance department with seeing a number of performance improvements in different areas but also a few possible regressions.

It has been a while since featuring a showdown of the Mozilla Firefox and Google Chrome web browsers on Linux. With some fresh benchmarks being overdue plus the new JetStream 3 browser benchmark having been announced last week, here is some fresh data for how these two dominant web browsers are competing on the modern Linux desktop from an Intel Panther Lake system running Ubuntu 26.04.

In testing thus far on Ubuntu 26.04, the KDE Plasma 6.6 desktop with the Wayland session is working pretty darn well and delivering a performance edge across many games/graphics workloads compared to the default GNOME 50 desktop. At least as far as AMD Radeon graphics are concerned, Plasma 6.6 is in quite impressive shape for the upcoming Ubuntu 26.04 LTS release.

With this week's release of Blender 5.1 I have begun benchmarking it on different CPUs and GPUs. In this article is an initial look at the positive impact Blender 5.1 is having on CPU-based rendering performance on Linux.

Last week I provided a look at the EXT4 and XFS performance from Linux 6.12 LTS through Linux 7.0 in its current development form. As mentioned in that article and as requested by many Phoronix readers, benchmarks have since wrapped up looking at how the Btrfs copy-on-write file-system performance has evolved since that late 2024 period and all major Linux kernel releases past that Long Term Support version.

Earlier this month were various Linux 7.0 file-system benchmarks showing how XFS is leading the race in the overall upstream Linux file-system performance on this forthcoming kernel. Stemming from that testing some premium supporters requested a fresh look at the historical performance of XFS as well as EXT4. So today's article is a look at how XFS and EXT4 have performed on every kernel release going back to Linux 6.12 LTS.

With a number of file-system improvements in Linux 6.19 and more file-system optimizations in Linux 7.0, it's past due for running some fresh file-system benchmarks. Here is a look at how the prominent file-system contenders are performing on the latest Linux 7.0 development kernel.

Announced this week at Mobile World Congress (MWC) by the Linux Foundation was the establishing of the OCUDU Ecosystem Foundation for advancing open-source AI-RAN (Radio Access Network) innovations. OCUDU is building a reference platform and innovations around 5G and early 6G network solutions. With OCUDU being benchmark-friendly, I have been putting the early code through some performance tests on current AMD EPYC and Intel Xeon server platforms.

With yesterday's stable release of the LLVM Clang 22 compiler it didn't take long for Phoronix readers to begin asking about the performance of this half-year feature update to this prominent open-source C/C++ compiler. What I am seeing so far are no big surprises with the performance largely being similar to Clang 21 across various open-source C/C++ workloads in the testing thus far. This initial round of reference benchmark results between LLVM Clang 22, Clang 21, and Clang 20 were done on an AMD EPYC Turin (Zen 5) Linux server.

While the version bump to 7.0 is driven solely by Linus Torvalds' versioning preferences, with Linux 7.0 there are many great changes to be found in this upcoming stable kernel version to power the likes of Ubuntu 26.04 LTS. Here is a recap of all the interesting changes with Linux 7.0.

When beginning some early Linux 7.0 kernel benchmarking this week for looking at its performance in its early development state, I started off testing on Core Ultra X7 "Panther Lake" in being hopeful for better performance with the maturing Arc B390 Xe3 graphics and the like. But I ended up finding Intel Panther Lake seeing some performance regressions on Linux 7.0. So next up I turned to an AMD EPYC Turin server since if regressions existed there at least it's much faster to carry out bisecting of the kernel performance regressions. But with that initial testing wrapped up, I didn't find any regressions like with Panther Lake and standing out were some rather enticing PostgreSQL database server performance benefits when running atop Linux 7.0.

With the Linux 7.0 merge window beginning to calm down ahead of the 7.0-rc1 release due out on Sunday, one of the areas I was most excited about benchmarking on Linux 7.0 was looking for any performance gains with the new Intel Core Ultra Series 3 "Panther Lake" given ongoing Intel Xe graphics driver improvements and other general kernel optimizations. Unfortunately, at large the Intel Panther Lake performance is moving in the wrong direction with the early Linux 7.0 benchmarking.

Back at the start of the Linux 6.19 kernel cycle I ran benchmarks showing some scheduler performance regressions with the new kernel. Fortunately, two weeks out from the Linux 6.19 stable release, merged this weekend was disabling the scheduler's NEXT_BUDDY feature due to performance regressions. Here are some fresh benchmarks looking at the latest Linux 6.19 Git state with/without NEXT_BUDDY and comparing it to Linux 6.18 stable for reference.

As part of my various end-of-year benchmarks, recently I looked at the Linux LTS kernel performance on AMD EPYC 9005 over the past year, the AMD EPYC Milan-X performance over the past four years, and various other performance comparisons over time to look the evolution of the Linux software performance. Another run I had carried out was looking at the AMD EPYC 8004 "Siena" series since its launch just over two years ago. Here is a look at how an up-to-date Linux software stack can deliver some additional performance gains for these energy efficiency and cost-optimized server processors.

When it comes to software leveraging Intel Advanced Matrix Extensions (AMX) functionality in modern Xeon processors, it's largely been limited to AI applications/libraries like oneDNN, OpenVINO, DeepRec, etc. But Intel now has another great open-source real-world AMX demonstrator with their Open Image Denoise library. This open-source library providing high quality denoising filters for images rendered using ray-tracing can end up benefiting big time from AMX-FP16 found with the newest Xeon 6 "Granite Rapids" processors. I ran some benchmarks of their new Open Image Denoise library with AMX-FP16 and was honestly blown away by the results.

When recently carrying out performance benchmarks of Intel Meteor Lake performance on Linux since launch day two years ago, the geo mean came in at 93% the original performance. Finding the performance trending clearly lower with an up-to-date Linux software stack compared to in December 2023 was quite surprising considering the rather nice gains we have seen over time on other Intel/AMD hardware. As noted in that article though, one of the possible explanations there is the Spectre BHI "Branch History Injection" vulnerability and microcode plus Linux kernel mitigations having come out post-launch and affecting Meteor Lake CPUs. Sure enough, follow-up tests looking at the Spectre BHI impact have revealed a measurable cost in a number of workloads for the Core Ultra processor.

It's been a while since running benchmarks of the Liquorix kernel as an enthusiast-tailored downstream version of the Linux kernel focused on responsiveness for gaming, audio/video production, and other creator/enthusiast workloads. In today's article is a look at how the latest Liquorix kernel derived from Linux 6.18 is competing against the upstream Linux 6.18 LTS kernel on the same system.

Over the past few years building the Linux kernel with Clang has matured a lot thanks to upstream improvements to both LLVM/Clang and the Linux kernel. As it's been a while since our last comparison for GCC vs. Clang built kernels on the resulting system performance, our latest year-end 2025 benchmarking is providing a fresh look at the Linux 6.19 upstream Git kernel built under the latest stable GCC 15 and LLVM Clang 21 compilers. Plus with the Clang-built kernel is also the option of the Link-Time Optimized (LTO) kernel for even greater performance.

As an end-of-year tradition at Phoronix for running a lot of year-over-year comparison performance benchmarks and other long-term performance evaluations, it's typically done on the higher-end hardware. That's done for a matter of time savings with maximum performance when running often 100~200+ benchmarks per article, the highest-end hardware typically being the most interesting in terms of features and capabilities, and more often than not getting flagship hardware review samples as opposed to the lower-end hardware. There have been benchmarks recently showing the big gains for AMD EPYC from a one year Linux LTS kernel upgrade, Intel Granite Rapids over the past year, and even the AMD Milan-X performance over the last four years, among other end-of-year 2025 articles. Today is a look at how the AMD Ryzen AI 5 "Krackan Point" CPU/iGPU performance has evolved simply over the last six months. It was a rather surprising twist how much better the Linux performance is over simply the past six months.

The recently released Linux 6.18 kernel is this year's Long Term Support version. As such it's sure to a see a lot of enterprise and hyperscaler uptake in being the annual LTS kernel version. While Linux 6.12 LTS will be maintained at least through the end of next year, upgrading to Linux 6.18 LTS can be very worthwhile from the performance perspective beyond the extended timeline until it will reach end-of-life. Here are benchmarks showing the performance advantages of upgrading from Linux 6.12 LTS to Linux 6.18 LTS for 5th Gen AMD EPYC "Turin" as well as an early look on the same server for the performance direction Linux 6.19 is bringing the kernel into 2026.

With Linux 6.19-rc1 released, the merge window for Linux 6.19 has now concluded. Here is a summary of the interesting Linux 6.19 new features and changes with this kernel version.

As the Linux 6.19 merge window winded down this weekend, I began running this development kernel on more systems. While there are some scheduler regressions currently with Linux 6.19 Git, for HPC workloads especially I am seeing some encouraging results using a flagship AMD EPYC 9965 2P server configuration.

Yesterday I noted some early performance regressions I've found on the Linux 6.19 kernel compared to Linux 6.18 LTS stable. Those initial benchmarks were on an AMD EPYC server. Since then I've seen many of the same workloads regressing similarly on an AMD Ryzen Threadripper workstation between Linux 6.18 and Linux 6.19 Git. Given the significant impact and AMD Threadripper processors always helping out to speed-up Linux kernel build times to make for a quicker and more manageable kernel bisecting experience, here is a look at some of the results for the Linux 6.19 performance regressions.

One of the areas I've been meaning to run more benchmarks on this season has been for the recently released Clang 21 compiler. Back in September when LLVM Clang 21 was debuting I ran some initial benchmarks and found it to deliver some nice performance gains on AMD EPYC Zen 5 but then have been busy with other benchmarks/articles for expanding that testing. Recently with having some spare cycles and gratis access still to the Microsoft Azure HBv5 instance for AMD EPYC Zen 4 with HBM3, I ran some Clang 20 vs. Clang 21 performance benchmarks there for those wondering about any performance benefits of this new compiler release on Zen 4.

Posted last month were new Linux kernel scheduler-related patches rewriting the MM CID management code. The main takeaway for end-users from this set of 19 Linux kernel patches from an Intel engineer was seeing 14~18% improvement in a PostgreSQL database benchmark but that more benchmarks were needed. Curiosity got the best of me and I recently tested these patches on an AMD EPYC server to seeing some very enticing results for this in-development code.