Intel News Archives

In addition to x86-simd-sort 3.0 being released for speedy AVX-512 sorting, Friday also brought the release of oneDNN 3.3 as the deep neural network library that is part of oneAPI and focused on helping developers build out deep learning applications.

Earlier this year Intel posted x86-simd-sort as a blazing fast sorting library that makes use of AVX-512. When the popular Numpy began using it they found up to 10~17x faster sorts for 16-bit to 64-bit data types. Today Intel software engineers released x86-simd-sort 3.0 and it also comes minutes after OpenJDK merged a modified version of this speeding sorting code into that reference JDK codebase.

Intel Linux graphics driver engineers are doing a bit of house keeping to the i915 DRM driver and removing some pre-production hardware bits as well as the incomplete Ponte Vecchio support.

In addition to Vulkan sparse support that works with the existing i915 kernel driver, another exciting open-source Intel Vulkan driver development this week is landing a transfer queue implementation for DG2/Alchemist GPUs. This in turn can help with hybrid graphics systems and other situations.

Motivated by a 25% performance degradation seen on an Intel Xeon Scalable dual socket server due to Indirect Branch Restricted Speculation (IBRS), Red Hat's Waiman Long has been working on a patch series to update the IBRS handling in different conditions for affected Intel processors on Linux.

The biggest hindrance for using Intel Arc Graphics for Linux gaming has been the lack of Vulkan sparse support as needed for running many newer Windows DirectX 12 games atop Valve's Steam Play with Proton using VKD3D-Proton. Intel recently did implement Vulkan sparse support for ANV in Mesa 23.3 but it only works with their yet-to-be-upstreamed and still-experimental Xe kernel driver. But now Intel Linux graphics driver engineers have managed to pull off a solution for getting the sparse resources supported while using the existing i915 kernel DRM driver.

Similar to Embree 4.0 adding Intel graphics acceleration earlier this year via SYCL, Intel's OpenVKL library that is also part of their oneAPI rendering toolkit has now added SYCL graphics acceleration with OpenVKL 2.0.

Intel's Linux graphics driver engineers have begun submitting their feature changes to DRM-Next of new i915 kernel driver feature material they are preparing for the Linux 6.7 cycle this winter.

With the Intel Arrow Lake NPU being very similar to Meteor Lake for this neural processing unit, the patches enabling that NPU for next-gen Intel Core CPUs was submitted as a "fix" for the ongoing Linux 6.6 cycle.

Intel engineers have merged to Mesa 23.3 the initial open-source "ANV" Vulkan driver support for sparse resources! This is the important feature needed for running a number of newer Direct3D 12 games with Steam Play (Proton) via VKD3D-Proton with Intel graphics hardware.

Earlier this month I noted that Intel's open-source Linux graphics driver engineers had begun working on the OpenGL and Vulkan Mesa driver support for Xe 2 graphics as to be found with Lunar Lake "LNL" processors. Since then the initial hardware enablement work has only continued heating up.

Intel's open-source software engineers are known for many great performance optimizations to the Linux kernel. Over the years Intel has contributed countless performance optimizations to the kernel and related Linux components that have made significant improvements not only for Intel hardware but x86_64 as a whole and at times CPU architecture independent improvements. One of their newest performance optimizing patch series is around Per-CPU Pageset (PCP) high auto-tuning.

Announced last year at the Intel Vision conference was the Habana Labs Gaudi2 and Greco AI hardware. Since then we've seen a lot of Linux kernel driver work happen for enabling the Gaudi2 second-generation training and inference AI processor while there hasn't been anything real in the way for Greco, which was the successor to the Goya AI processor. Now references to Habana Labs Greco are being removed from the driver.

Since announcing the Advanced Performance Extensions (APX) and AVX10 back in July, Intel's open-source compiler engineers have been busy preparing the GCC and LLVM/Clang compiler toolchains for these major CPU extensions to be found with future Intel processors.

With the Linux 6.7 kernel this winter there is a new feature coming to Intel's QuickAssist Technology (QAT) device driver that will allow more efficient use with QAT Gen4 hardware such as the latest-generation Intel Xeon Scalable "Sapphire Rapids" processors.

While Intel divested its storage business and Intel Optane was sadly discontinued, one of the interesting open-source software projects from its storage efforts has been DAOS, the Distributed Asynchronous Object Storage engine. Version 2.4 of the DAOS software-defined object store designed for high-speed storage was released this past week.

With Meteor Lake comes the introduction of the Versatile Processing Unit (VPU) that is now marketed by Intel as the Neural Processing Unit (NPU). Recent versions of the Linux kernel have the "IVPU" accelerator driver to support Meteor Lake's VPU/NPU while now a patch is pending to extend that support for next-generation Arrow Lake processors.

Intel engineers have published their Compoute Runtime 23.30.26918.9 that provides their open-source Level Zero and OpenCL support for use on Windows and Linux platforms with Intel integrated/discrete graphics hardware.

Intel announced this morning on the second day of their Innovation 2023 conference that they are collaborating with software vendors such as Red Hat, Canonical, and SUSE for providing Intel-optimized Linux distributions.

Last year Intel announced Project Amber as an effort to verify the trustworthiness of clouds. Project Amber was talked up as "an innovative service-based security implementation" for the remote verification of the trustworthiness of compute assets. Project Amber is now rolling out as the Intel Trust Authority.

Intel is kicking off their Innovation 2023 conference in San Jose with many exciting announcements. Freed from embargo this morning is news around their upcoming mobile and server processors, lots of AI talk, Intel's continued software advancements to complement their hardware, and more.

Stemming from Intel engineers finding significant overhead in some Linux scheduler functions when running PostgreSQL within a Docker instance, a new scheduler patch is on the way for Linux 6.7 that will help out at least Ice Lake and Sapphire Rapids with some migration-heavy workloads. With the change being in the common scheduler code, it's also likely to help out other hardware platforms too.

Since 2021 the Itanium IA-64 code was orphaned in the Linux kernel and over the course of this year there's been talk of retiring the Itanium code from the kernel, a.k.a. strip it out. It looks like 2023 will end with the Itanium IA-64 code indeed being removed from the Linux kernel.

Intel's OpenVINO 2023.1 was just published to GitHub as the newest version of this open-source toolkit for optimizing and deploying AI workloads across their CPUs, GPUs, and now also having official support for the new VPU being found with Meteor Lake SoCs.

Tucked away within a merge to Mesa entitled, "intel/genxml: Support importing, which cuts the overall xml file size almost in half" Intel Linux graphics driver engineers are slowly beginning to start work on the "Xe2" graphics support to be found with Lunar Lake processors.

One of the nifty new features with Intel Xeon Scalable "Sapphire Rapids" processors not talked about much is the In-Field Scan (IFS) functionality for allowing various tests to be carried out prior to deploy the new processors or over time in the data center for trying to spot any silicon-level issues. IFS aims to uncover any processor issues not caught by ECC, parity checks, RAS, or other safeguards when new CPUs reach the data center or as they age over time.

Intel's open-source Compute Runtime 23.26.26690.22 was released today, which is a big update for this OpenCL and oneAPI Level Zero (L0) stack for Windows and Linux systems. Due to the summer holidays and Intel's current release regiment for the Compute Runtime, v23.26.26690.22 is the first new release since mid-July.

Intel today announced Thunderbolt 5 as their next-gen Thunderbolt standard that will allow 80 Gbps of bi-directional bandwidth or a "Bandwidth Boost" mode of up to 120 Gbps.

Intel's open-source "ANV" Vulkan driver used by Linux systems is now enabling by default the VK_EXT_mesh_shader extension for their graphics processors with mesh shader support.

One of the driver additions we've been eager to see for the mainline Linux kernel that didn't pan out in time for the recently closed Linux 6.6 merge window is the Intel Xe DRM kernel graphics driver as a modern alternative to their i915 driver. The Xe driver better supports non-x86 CPU architectures, better designed and more performant around their modern integrated and discrete GPUs, and overall is able to make better design choices and improvements in being a clean sheet driver design compared to all the code that has built up in i915 over the years. But for getting the Xe driver upstream even in experimental form, first some necessary DRM scheduler patches need to be ironed out.

Sub-NUMA Clustering with Intel Xeon processors allows for splitting up the CPU cores, cache, and memory into multiple NUMA domains for enhancing the performance of NUMA-aware applications. While SNC can help in a number of cases especially plenty of HPC and server workloads, currently it's not properly supported if making use of Resource Director Technology (RDT) on modern Intel CPUs. That is in the process of changing with new Linux kernel patches being worked on by Intel.

As a big win for Intel's QuickAssist Technology (QAT) accelerator found as an option with Sapphire Rapids processors and prior QAT hardware, there's been a QAT Zstd plug-in to provide big performance/efficiency benefits. Version 0.1 of that plug-in was released today.

Intel's open-source "ANV" Vulkan Linux driver has landed a set of 15 patches that can further help the Linux gaming performance for those using DG2/Alchemist Arc Graphics.

Greg Kroah-Hartman last Friday sent out the "char/misc" changes for the Linux 6.6 kernel merge window that include the catch-all of different driver changes. This cycle the char/misc updates aren't particularly exciting but the Intel PECI updates do stand-out.

Ahead of Intel expecting to formally reveal more Meteor Lake processor details in the coming weeks, Intel's open-source Linux graphics driver developers are ready to declare the new integrated graphics stable with the upstream Linux kernel -- dropping the existing flag that treated them as experimental and not enabled by default unless passing a special kernel module parameter.

Following Intel's acquisition of Tower Semiconductor falling through due to failing to obtain all the necessary regulatory approvals, Intel and Tower today announced a foundry agreement.

The decision last month for the Linux kernel to disable random number generation (RNG) for all AMD fTPMs ended up having some unintended consequences on Intel systems that ended up breaking S3 suspend behavior.

Sriram Ramkrishna at Intel, who serves as the community manager and developer relations for oneAPI, held a virtual oneAPI meetup this week with Red Hat's David Airlie. Airlie should not need any introduction for longtime Phoronix readers given his longtime contributions to the Linux kernel graphics drivers, Mesa, and related open-source graphics work at Red Hat. Airlie shared some interesting remarks around the current Linux GPU compute stacks from the different vendors and associated challenges.

The perf event changes were submitted this week for the ongoing Linux 6.6 kernel merge window with changes this cycle for Intel, AMD, and Arm.

The media subsystem updates were sent out today for the Linux 6.6 kernel and most notably is introducing the Intel IVSC MEI drivers as well as extending the Intel IPU bridge logic to work with these new drivers.

The Intel Shadow Stack support that is part of their Control-flow Enforcement Technology (CET) has finally been merged for the Linux 6.6 kernel after it was previously rejected by Linus Torvalds.

The Linux mitigation for the Intel Gather Data Sampling (GDS) "Downfall" vulnerability was updated to reflect all Skylake and Kabylake CPUs being vulnerable to this nasty issue. Due to those Skylake client processors reaching the end of their official support life at Intel, the original Linux mitigation for GDS/Downfall didn't properly protect those older Core processor models.

The Intel-led SVT-AV1 open-source AV1 video encoder is out with a major release that delivers on more performance improvements across the board.

Seemingly for Intel's upcoming Raptor Lake Refresh processors, some new graphics PCI IDs were added today to Mesa for the Iris Gallium3D and ANV Vulkan drivers.

In addition to AMD sending out DCN 3.5 display patches for that next-gen display IP block presumably for their upcoming Ryzen 8000 series APUs, Intel's open-source engineers today sent out the patches enabling Lunar Lake display support for their i915 kernel driver while there is also support baking for their in-development Xe kernel driver.

Intel is well regarded for their vast open-source contributions from being a major contributor to the Linux kernel and other areas like Mesa, GCC/glibc, and other key open-source projects to various niche projects like ConnMan and other smaller software projects. Debuting a few months ago as one of the newest open-source Intel projects catching us by surprise was Intel One Mono as a font designed for developers. Today brings a new version of that font.

Intel ISPC 1.21 is now available as the newest feature update to this Implicit SPMD Program Compiler for a C language variant focused on single program, multiple data (SPMD) programming for exploiting the full potential of modern CPUs and GPUs.

Intel engineer Ilpo Järvinen posted a set of Linux kernel driver patches to introduce a new "bwctrl" PCI Express Bandwidth Controller driver and associated PCIe cooling driver to allow for limiting the PCIe link speed in the event of any system thermal issues.

While Intel has maintained the QATzip open-source compression library for demonstrating data compression using QuickAssist Technology (QAT) with DEFLATE/LZ4/LZ4s, Intel has also been working on QAT'ed Zstd for achieving some sizable victories in performance and power efficiency.

It was less than one month ago that Intel announced AVX10 as the successor to AVX-512. In that time Intel engineers have begun posting AVX10.1 enablement patches for GCC as well as beginning AVX10 discussions for the LLVM compiler stack. Overnight already the initial AVX10.1 enablement code has been merged into the GNU Compiler Collection.