NVIDIA News Archives

NVIDIA has launched their latest TITAN ultra high-end graphics card, the TITAN V, which they say is the most powerful PC GPU ever created. The TITAN V is based upon their Volta architecture and will set you back $2,999 USD.

With a NVIDIA Linux developer having confirmed a current driver performance regression affecting driver releases since the 378 series and not being worked around until the yet-to-be-released 390.xx beta driver, I decided to carry out some tests.

Following NVIDIA's call for feedback on their effort to create a new device memory allocator API that would be of equal use to the upstream open-source drivers and potentially replace (or indirectly used by) the Wayland compositors in place of the existing GBM API and NVIDIA's failed EGLStreams Wayland push, their next steps continue to be formulated.

If you think recent NVIDIA Linux driver releases have been slowing down your games, you are not alone, especially if you are running with a GeForce graphics card having a more conservative vRAM capacity by today's standards.

For those using the NVIDIA 387 "short-lived" driver series, the 387.34 release is now available with just three changes noted.

If you are using the NVIDIA proprietary graphics driver and anxious to try out the Linux 4.15 kernel for its many new features/improvements, unfortunately you will need to wait a few days as the current public driver is broken against this latest code.

After apologizing how they handled the EGLStreams proposal for NVIDIA Wayland support, James Jones of NVIDIA is trying to get the development of their proposed generic device memory allocator library back on track.

A few days back I wrote about FFmpeg picking up NVDEC-accelerated H.264 video decoding and since then more FFmpeg improvements have landed.

The NVIDIA-owned PGI has announced their latest monthly update to their proprietary CPU/GPU compiler stack for Windows, Linux, and macOS systems.

Besides the NVIDIA 387.22 Linux driver update released earlier this week, today they are issuing the 384.98 update in their long-lived driver branch.

The NVIDIA GeForce GTX 1070 Ti has begun shipping.

NVIDIA has shipped the 387.22 Linux driver today as their first stable release in the 387.xx series.

Today's GeForce GTX 1070 Ti announcement is hardly a surprise given all the recent leaks about this new card, but NVIDIA announced it today and will be shipping in early November.

While the Jetson TX2 has been out since this past March and it's a phenomenal ARM development board, sadly the Direct Rendering Manager (DRM) driver support for it still isn't ready with the mainline Linux kernel.

NVIDIA has today released an updated version of their Linux Graphics Debugger to help game/application developers in analyzing issues and performance problems around OpenGL 4.x on GeForce/Quadro GPUs.

A video decoder driver for the NVIDIA Tegra is closer to the mainline kernel, but is focused on the (older) Tegra 2 chips.

NVIDIA has launched their first beta driver in the 387 series for Linux.

NVIDIA has announced a new open-source project: NVDLA.

Numba is designed to allow for high performance Python JIT-compiled code designing for C/C++ levels of performance while using LLVM for optimizations and allowing GPU offloading too. NVIDIA is promoting Numba in the context of CUDA.

While NVIDIA's Vulkan Linux driver performance has already been very good, it's potentially even better now with the newest NVIDIA Vulkan beta driver.

NVIDIA has kicked off their GTC world tour and in beginning that, they have launched the CUDA 9.0 toolkit.

While NVIDIA isn't doing much to help out Nouveau, at least the company is contributing to the open-source Linux graphics ecosystem in other ways. In addition to presenting at XDC2017 this week on the Unix device memory allocator API and DeepColor / HDR support, they also presented on server-side GLVND.

One day after releasing updated GeForce Linux legacy drivers, NVIDIA is now out with an update to their long-lived 384 branch.

Besides working on the new Unix device memory allocator project, they have also been engaged with upstream open-source Linux developers over preparing the Linux desktop for HDR display support.

James Jones of NVIDIA presented this morning at XDC2017 with their annual update on a new Unix device memory allocation library. As a reminder, this library originated from NVIDIA's concerns over the Generic Buffer Manager (GBM) currently used by Wayland compositors not being suitable for use with their driver's architecture and then the other driver developers not being interested in switching to EGLStreams, NVIDIA's original push for supporting Wayland.

NVIDIA has issued new releases of its two legacy drivers for Linux.

For those wanting the bleeding-edge NVIDIA Vulkan driver support, a new beta was pushed out today providing same-day support for the Vulkan 1.0.61 update.

For those looking for a very capable ARM developer board but have previously been put off by the Jetson TX1 at $579 USD, they now have a $199 developer board.

NVIDIA today released the 384.69 Linux driver as their latest release in the 384 "long-lived" series.

NVIDIA is working on a new OpenGL memory usage reporting extension, NV_query_resource. Before anyone jumps though to bash NVIDIA over coming up with yet-another-memory-reporting extension for OpenGL, this one is aimed at reporting the usage at an object-level rather than just overall amounts.

NVIDIA's driver team has today released new Vulkan beta drivers for both Windows and Linux.

Just a quick PSA for NVIDIA binary Linux driver users wishing to play with the latest development kernel...

NVIDIA is ending out the week with their first release candidate of CUDA 9.

For NVIDIA Linux users read our OpenGL 4.6 overview if you haven't already and then go forth and download the new experimental driver.

NVIDIA's Unix graphics team has today released some updated binary display drivers.

For those with NVIDIA Jetson TX1/TX2 developer boards, the JetPack 3.1 software bundle is now available with NVIDIA's latest compute bits paired with the Linux 4.4 LTS kernel.

NVIDIA and Carnegie Mellon University have developed a new open-source shading language in step with a new compiler framework.

NVIDIA has once again managed a same-day driver update for matching a new Vulkan release.

At the end of June I posted some Vulkan vs. OpenGL Linux Game CPU Core Scaling using RADV/RadeonSI with a Polaris graphics card. At that time I also carried out some NVIDIA CPU core scaling results in a Vulkan vs. OpenGL manner, but simply forgot to post those numbers until now.

A few days back NVIDIA released the Vulkan 381.10.10 Linux beta that featured performance improvements as well as new Vulkan/OpenGL interoperability extensions. That beta driver has already been succeeded by a new driver.

Today NVIDIA released their first 384 series Linux driver beta and for the occasion I fired up some fresh OpenCL / Vulkan / OpenGL benchmarks in seeing if there are any performance changes for users to see with this new series that will eventually succeed the 381.22 stable release.

NVIDIA has today announced the 384.47 beta driver for Linux, which succeeds their current 381 short-lived stable release series.

With NVIDIA just releasing a new beta Vulkan driver that in addition to having new Vulkan extensions and better Vulkan/OpenGL interoperability also has "various performance improvements", I couldn't resist running some benchmarks.

NVIDIA quietly released a new Vulkan beta driver that offers up a number of new OpenGL and Vulkan extensions that haven't yet received widespread exposure. NVIDIA Vulkan users, meet the 381.10.10 release.

NVIDIA has made their TensorRT 2 library publicly available today as the newest major update to their deep-learning inference optimizer and run-time.

NVIDIA has today issued a new public Vulkan beta driver for Windows and Linux.

In CPU-bound Linux games, the NVIDIA Linux driver still appears to perform better than the newest RadeonSI Gallium3D code.

For those looking for a low-profile, single-slot graphics card for an HTPC box or so, more of them should be hitting the market in the form of NVIDIA's new GeForce GT 1030.

NVIDIA at their annual GPU Technology Conference (GTC'17) have provided more public details about the forthcoming CUDA 9 compute update.

NVIDIA CEO Jen-Hsun Huang used the company's GPU Technology Conference to today announce the Volta V100 accelerator for data centers and HPC workloads.