NVIDIA News Archives

Ahead of NVIDIA Vera ramping up, the upcoming Linux 7.2 kernel is adding the ACPI CPPC v4 support authored by a NVIDIA engineer.

Last month CUDA-Oxide was introduced as an experimental Rust-to-CUDA compiler. From pure Rust programming language code, one can write CUDA GPU kernels in a "safe(ish)" manner with the CUDA-Oxide compiler emitting NVIDIA PTX output directly. Out today is the second update to CUDA-Oxide.

EGLStreams was NVIDIA's original route to supporting Wayland with their official Linux graphics driver stack. Adoption was limited and driver vendors outside of NVIDIA didn't end up going with EGLStreams/EGLDevice. Thankfully, NVIDIA corrected course long ago with DMA-BUF, GBM, and KMS support that aligns with the rest of the ecosystem, and now that old code path is being removed from GNOME Mutter.

While the upstream, open-source Nouveau driver already supports NVIDIA's Hopper and Blackwell graphics processors with the GPU System Processor (GSP) code path, the bring-up of the Rust-written Nova driver remains ongoing. Out this week is the 12th iteration of the Hopper and Blackwell enablement for this future open-source NVIDIA Linux driver.

An open-source developer has created NBD-VRAM as a way to create swap space on your consumer NVIDIA GPU's video memory under Linux.

Jensen Huang used his Computex keynote today to formally announce RTX Spark as their new superchip for compact desktop PCs and laptops.

NVIDIA on Tuesday released CUDA 13.3 as another significant advancement for their unified GPU programming stack for NVIDIA hardware.

NVIDIA is kicking off the new week with their first Linux driver beta in the R610 driver series that is succeeding the current R595 release branch.

DXVK-NVAPI 0.9.2 is now available for this implementation of NVIDIA's NVAPI/NVOFAPI interfaces atop DXVK and VKD3D-Proton that is used in turn by Valve's Steam Play (Proton) for enhanced NVIDIA Linux gaming support.

The open-source, community-developed NVIDIA-VAAPI-Driver that provides a Video Acceleration API (VA-API) implementation built atop NVIDIA's NVDEC video decode interface is out with a new release. This is the open-source project that's motivated by getting accelerated video decoding to work within Mozilla Firefox and other apps when running with NVIDIA's packaged Linux driver.

A new NVIDIA Labs project is greatly improving the capabilities of using the Rust programming language for developing CUDA kernels for NVIDIA GPUs.

NVIDIA compiler engineers are looking to develop a standalone tool that could be upstreamed into the GNU Compiler Collection (GCC) codebase for generating AutoFDO profiles for consumption by GCC in turn for better benefiting from automatic feedback directed optimizations (FDO) in the name of better performance.

NVIDIA on Friday released the 595.44.06 beta driver build as their newest Vulkan developer beta for Linux. This was joined by the NVIDIA 595.46 Windows Vulkan beta and there are performance improvements in tow and more work on their descriptor heaps support.

The legacy xf86-video-nv driver for user-space mode-setting on old NVIDIA GPUs is out with a rather rare release and the first in over two years with a collection of different bug fixes.

Last month we showcased GreenBoost as an open-source means of augmenting NVIDIA GPU vRAM with system RAM and NVMe storage. This memory tiering solution for NVIDIA GPUs was developed by an open-source developer with a focus on CUDA and allowing larger LLMs to be handled on graphics cards with smaller vRAM capacities. There was a setback to the project due to NVIDIA legal but now the project is going in new form and also has introduced GreenBoost-Proton for helping Linux gaming on NVIDIA hardware.

Last year NVIDIA announced the new CUDA Tile programming model as one of the biggest updates ever to the CUDA platform. CUDA Tile brings a virtual ISA for tile-based parallel programming and they subsequently open-sourced the CUDA Tile IR as an intermediate representation built atop LLVM's MLIR. Now they are looking to hire additional LLVM compiler engineers to help foster their CUDA Tile initiatives.

Following the DRM Color Pipeline API making it into the Linux 6.19 kernel, NVIDIA today released a preview Linux driver with their support for the DRM per-plane color pipeline API that will benefit the broader Linux/Wayland desktop HDR ambitions.

An important set of Linux scheduler patches were posted for review on Thursday for improving the SMT-aware asymmetric CPU capacity handling. These patches to improve the Linux kernel scheduler around CPU Simultaneous Multi-Threading (SMT) is needed after NVIDIA engineers discovered up to a ~2x performance drop for CPU-intensive workloads on their upcoming Vera Rubin platform.

Building off the NVIDIA 595.45.04 Linux beta driver that brought DRI3 v1.2 support and new Vulkan capabilities, the NVIDIA 595.58.03 Linux driver released this morning as the first stable Linux driver build in the R595 release branch.

KubeCon Europe is running this week in Amsterdam and NVIDIA used the event to talk up their open-source work around AI and newest open-source contributions.

An interesting Linux detail from today's NVIDIA GTC 2026 kickoff is that Canonical will be integrating NVIDIA's DOCA-OFED software framework into the Ubuntu Linux archive for leveraging the high-speed networking stack for HPC and AI.

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. The intent here with GreenBoost is a CUDA caching layer to more easily run larger AI models for LLMs that otherwise won't fit solely in your graphics card's dedicated vRAM.

Less than a week after their prior NVIDIA Vulkan beta driver in the R595 driver series, today NVIDIA released another new Vulkan beta driver for Linux and Windows systems.

Last week NVIDIA released the 595.45.04 Linux driver beta as their first release in the R595 series for Linux and it's running very well in initial testing. Today as part of their Vulkan developer beta program, they have released the NVIDIA 595.44.02 driver that brings some new Vulkan API features.

With CUDA 13.2 that is now shipping, NVIDIA has provided official support for Red Hat Enterprise Linux compatible distributions/downstreams like AlmaLinux to CUDA. With this official NVIDIA CUDA support for these RHEL-compatible distributions, NVIDIA is also allowing the NVIDIA packages to be distributed directly from the OS package repositories.

Following the recent NVIDIA R595 driver release for Windows, NVIDIA today released the 595.45.04 driver for Linux users as a beta version in the R595 release stream.

One of the latest NVIDIA open-source contributions this week wasn't for the in-development Nova kernel driver but for enhancing the existing Nouveau kernel driver. The patch posted is for bringing up the NVIDIA GA100 GPU under Nouveau using the GPU System Processor (GSP).

A few months ago at SIGGRAPH was a demo of Blender with NVIDIA Deep Learning Super Sampling (DLSS) integration. The pull request is now open for landing NVIDIA DLSS support into Blender for better quality upscaling/denoising and performance but concerns persist over the licensing due to NVIDIA DLSS binaries.

The NVIDIA-VAAPI-Driver 0.0.15 was released overnight as this VA-API driver implementation built atop NVIDIA's NVDEC interface used by their proprietary user-space driver stack. The purpose of NVIDIA-VAAPI-Driver as this community open-source project continues to be around enabling video acceleration for NVIDIA GPUs with the Firefox web browser on Linux that supports the VA-API interface but not NVIDIA's NVDEC.

As a wonderful New Year surprise, there's good momentum on NVIDIA graphics support for the BeOS-inspired Haiku open-source operating system.

NVIDIA's Olympus are the ARM64 cores found within the upcoming Vera CPU that will be paired with Rubin. Olympus cores are claimed to be twice as fast as NVIDIA's current CPU cores found in Grace and based on Neoverse-V2. Earlier this year the open-source compilers landed initial support for Olympus while now a proper CPU scheduling model has been upstreamed into LLVM 22.

This year there was a lot of going on in the NVIDIA Linux world from their official driver stack seeing better Wayland support to a lot on the open-source scene from NVIDIA engineers contributing a lot directly to the Rust-based Nova open-source driver that continues taking shape, the Mesa NVK Vulkan driver becoming more performant and capable, and a lot of other happenings. Here is a look back at the most popular NVIDIA content of 2025 on Phoronix.

As a wonderful Christmas gift to open-source fans, NVIDIA dropped their proprietary license on the CUDA Tile intermediate representation and has now made the IR open-source software.

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. It's an interesting read/watch around the safety of Linux (or not) for such strict safety environments.

NVIDIA has promoted their R590 driver series to stable with the release today of the NVIDIA 590.48.01 Linux driver as their latest new feature branch version.

NVIDIA just released CUDA 13.1 for what they claim is "the largest and most comprehensive update to the CUDA platform since it was invented two decades ago." The most notable addition with the CUDA 13.1 release is CUDA Tile as a new tile-based programming model.

NVIDIA today released the 590.44.01 Linux driver build as the first beta of their R590 series driver branch for Linux customers.

NVIDIA has a number of Linux kernel patches on the way to the Linux 6.19 kernel in preparing for 1.6 Tb/s networking on NVIDIA-Mellanox hardware.

NVIDIA today issued the 580.94.11 driver release as their newest Vulkan beta driver for Linux customers. Most notable with this beta driver update is adding VK_EXT_hdr_metadata support.

In addition to showing the need for unifying DRM driver-side APIs within the Linux kernel, NVIDIA's Linux graphics driver team at XDC2025 also showcased the shortcomings of screencasting under Wayland.

NVIDIA engineers continue working a lot on the in-development and in-tree open-source Nova kernel driver for their GPUs. Sent out on Friday night were the Turing enablement patches for this Rust-written Nova-Core driver code.

CUDA 13.0 Update 2 is now available as the latest incremental improvement to NVIDIA's compute stack.

NVIDIA engineers continue working a lot on the open-source and upstream Nova driver for the Linux kernel. This modern, Rust-written open-source NVIDIA driver is still taking shape as an alternative to NVIDIA's official downstream open-source driver and the aging and reverse-engineered Nouveau driver. Out on the horizon for Nova is Hopper and Blackwell GPU support.

NVIDIA is taking the open-source and upstream "Nova" kernel graphics driver quite seriously for their hardware. Hitting the mailing lists on Friday night were initial patches in beginning to make preparations toward "next-gen GPU" support. Digging into the comments, it's indeed for post-Blackwell GPUs.

Just over one year ago NVIDIA posted open-source Linux GPU driver code for GPU virtualization "vGPU" support. That NVIDIA vGPU driver work was recently revised while still posted under a request for comments (RFC) flag.

Debuting today is the newest NVIDIA 580 Linux driver series release.

Following AMD announcing the end of the AMDVLK Vulkan driver development in favor of focusing on the Mesa RADV driver for Linux systems, Red Hat engineer David Airlie who was one of the co-lead developers of the RADV driver shared some interesting insight on NVK as the open-source NVIDIA Vulkan driver being developed within Mesa.

SUSE in partnership with NVIDIA today announced making the NVIDIA CUDA TOolkit officially available on all SUSE platforms.

Released just over one month ago was the general availability of CUDA 13.0 while out this week is CUDA 13.0 Update 1 as the first incremental step forward to CUDA 13.

As part of the Rust DRM drivers now having their own development tree, sent out today was the first pull request from the drm-rust-fixes branch.