Enlarge / An AI-generated image of the White House in front of a cybernetic background. (credit: Midjourney)

On Thursday, the White House announced a surprising collaboration between top AI developers, including OpenAI, Google, Antrhopic, Hugging Face, Microsoft, Nvidia, and Stability AI, to participate in a public evaluation of their generative AI systems at DEF CON 31 , a hacker convention taking place in Las Vegas in August. The event will be hosted by AI Village , a community of AI hackers.

Since last year, large language models (LLMs) such as ChatGPT have become a popular way to accelerate writing and communications tasks, but officials recognize that they also come with inherent risks. Issues such as confabulations , jailbreaks, and biases pose challenges for security professionals and the public. That's why the White House Office of Science, Technology, and Policy endorses pushing these new generative AI models to their limits.

"This independent exercise will provide critical information to researchers and the public about the impacts of these models and will enable AI companies and developers to take steps to fix issues found in those models," says a statement from the White House, which says the event aligns with the Biden administration's AI Bill of Rights and the National Institute of Standards and Technology's AI Risk Management Framework .

Enlarge / I wish I had more games installed for iPad-on-the-couch photo purposes, but I just don't keep that many games on my drive at once! (credit: Kevin Purdy)

Nvidia's GameStream had one job, the one in its name: stream games from the Nvidia graphics card inside your PC to the Nvidia Shield hooked up to your TV (or, back in the day, a Shield tablet ). It did this job fairly well, making setup simple and optimizing games with some custom stream-smoothing. Now Nvidia is removing GameStream from Shield devices —but an even better DIY game-streaming solution is already available. Let's take a look at it and talk to the developers about why and how they made it.

Nvidia's GameStream was a key part of the hardware company's pitch for its Shield products. (credit: Nvidia)

Nvidia is done with local streaming

Nvidia says a Shield update arriving this month will make it so "the GameStream feature will no longer be available in app." If you try to skip the Shield update, GameStream will still stop working at some point (and possibly be removed from the GeForce Experience app in Windows). In the meantime, trying to dodge that update means not using GeForce Now , one of Nvidia's recommended replacements, on your Shield and missing out on all the other update fixes and features that arrive with system updates.

If you're a Shield owner, like I am, this stinks. Shield devices have merits of their own , receiving the longest and most consistent stream of updates of any Android/Google TV device ever released. They're still perfectly functional as stream boxes (and even more appealing if Google lands an NFL package ). But a big benefit of having both a Shield and a GeForce graphics card will soon be shunted.

Enlarge / Nvidia's GeForce RTX 4080. (credit: Andrew Cunningham)

The video encoding hardware built into GeForce GPUs is getting a small boost, according to a quietly updated Nvidia support page (as spotted by Tom's Hardware ). Previously, the NVENC encoder built into GeForce GPUs could encode up to three video streams simultaneously. Now, most GPUs supported by Nvidia's current drivers can encode up to five streams of video simultaneously, unlocking capabilities that had always been present in the hardware but that were software-limited in consumer GPUs.

It's unclear exactly when Nvidia made this change, but archival snapshots on the Internet Wayback Machine show the old three-stream limit as recently as March 18, so you may need to install the most recent drivers to unlock the additional encoding capabilities. Your video quality settings may also limit the number of video streams you can encode simultaneously.

Most GeForce GPUs going back to the 2014-era Maxwell architecture now support the extra simultaneous streams, so you don't need a new or powerful video card to benefit from the change (though there are some models, particularly MX-series GPUs for budget laptops, that still don't have any video encoding capabilities, presumably because they're missing the hardware). Models as old as the GeForce 750 Ti are on the list, as are most GeForce 900, 1000, 2000, 3000, and 4000-series cards. The kinds of video you can encode will still come down to what your GPU's hardware encoder actually supports; that Nvidia support document lists supported codecs, color depths, and other specs for each GPU.

Enlarge / Nvidia's GeForce RTX 4080. (credit: Andrew Cunningham)

Nvidia released a new driver update for its GeForce graphics cards that, among other things, introduced a new Video Super Resolution upscaling technology that could make low-resolution videos look better on high-resolution screens. But the driver (version 531.18) also apparently came with a bug that caused high CPU usage on some PCs after running and then closing a game.

Nvidia has released a driver hotfix (version 531.26) that acknowledges and should fix the issue, which was apparently being caused by an undisclosed bug in the "Nvidia Container," a process that exists mostly to contain other processes that come with Nvidia's drivers. It also fixes a "random bugcheck" issue that may affect some older laptops with GeForce 1000-series or MX250 and MX350 GPUs.

Not all PCs running the newer Nvidia drivers were being affected by the bug— some reporters observed the behavior on their systems, while others didn't . Even relatively low CPU usage in the 10 to 15 percent range can have a noticeable performance impact, taking CPU cycles from other tasks and preventing the CPU from going into an idle state. This generates more heat and uses more power and could also affect the battery life of laptops.

Enlarge / Currently, only 30- and 40-series GPUs are supported. (credit: Nvidia)

Nvidia's latest GPU driver introduces its new AI-based upscaling technique for making lower-resolution videos streamed offline look better on a high-resolution display. Now available via the GeForce driver 541.18 released on Tuesday, Nvidia's RTX Video Super Resolution (VSR) successfully cleaned up some of the edges and blockiness of a 480p and 1080p video I watched on Chrome using a 3080 Ti laptop GPU-powered system, but there are caveats.

By Nvidia's measures , 90 percent of video streamed off the Internet, be it from Netflix, YouTube, Hulu, Twitch, or elsewhere, is 1080p resolution or lower. For many users, especially those with Nvidia GPU-equipped systems, when moving to 1440p and 4K screens, browsers upscale this content, which can result in image artifacts like soft edges.

Nvidia VSR, which (somehow) shouldn't be confused with AMD VSR ( Virtual Super Resolution, targeting lower-resolution displays), uses the AI and RTX Tensor cores in Nvidia's 30- and 40-series desktop and mobile GPUs to boost sharpness and eliminate "blocky compression artifacts" when upscaling content to 4K resolution, per a blog post Tuesday by Brian Choi, Nvidia's Shield TV product manager.

Enlarge (credit: Arm)

Brussels is set to launch a formal competition probe early next month into Nvidia’s planned $54 billion takeover of British chip designer Arm, after months of informal discussions between regulators and the US chip company.

The investigation is likely to begin after Nvidia officially notifies the European Commission of its plan to acquire Arm, with the US chipmaker planning to make its submission in the week starting September 6, according to two people with direct knowledge of the process. They added that the date might yet change, however.

Brussels’ investigation would come after the UK’s Competition and Markets Authority said its initial assessment of the deal suggested there were “serious competition concerns” and that a set of remedies suggested by Nvidia would not be sufficient to address them.

Read 6 remaining paragraphs | Comments

After releasing upgrading their stable drivers with version 460.80 following the release of the the RTX 3050 and RTX 3050 Ti for laptops - a new Vulkan Beta Driver is out now.

Vulkan Beta 455.50.19 is a pretty small one, here's what's changed:

• New:
  • VK_EXT_provoking_vertex
• Fixes:
  • Added support for Vulkan provisional video session picture parameters objects

Reminder: This special Vulkan beta driver is where all the shiny new stuff goes in before making its way into the stable release for everyone. Really, it's mostly aimed at developers and serious enthusiasts. Unless you need what's in them, it's generally best to use the stable drivers.

The newest stable versions of the main NVIDIA driver for Linux are at 460.80 released on May 11, 2021 from their "Production Branch" series or 465.27 released on April 29, 2020 from their "New Feature" series. Confused? It's a lot of numbers to remember but both Production and New Feature series are fine for normal use.

Note: you should probably make sure your drivers are up to date, especially after the recent security issues .

See the driver page for more .

Following on from the news earlier that NVIDIA has released the RTX 3050 and RTX 3050 Ti for laptops , their new stable driver 460.80 is out for Linux.

As always for NVIDIA, they push out a new driver right away to get Linux support for their GPUs hooked up day and date with the release. The 460.80 driver adds support for:

• GeForce RTX 3050 Ti Laptop GPU
• GeForce RTX 3050 Laptop GPU
• T600 Laptop GPU
• T1200 Laptop GPU
• RTX A5000 Laptop GPU
• RTX A4000 Laptop GPU
• RTX A3000 Laptop GPU
• RTX A2000 Laptop GPU

It also includes some bug fixes:

• Fixed a bug that could cause AddressSanitizer to report a heap-buffer-overflow during initialization of the OpenGL and Vulkan libraries.
• Fixed a bug that could prevent a system from resuming from suspend when DisplayPort activity occurred while the system was suspended.
• Fixed a regression that prevented eglQueryDevicesEXT from correctly enumerating GPUs on systems with multiple GPUs where access to the GPU device files was restricted for some GPUs.
• Fixed a regression that could cause system hangs when changing display resolution on SLI Mosaic configurations.
• Fixed a bug that could result in blank displays when driving multiple displays at the same resolution using active DisplayPort dongles.

This is part of their "Production Branch", so it's good for everyone to upgrade. See the driver release page .