Since no one else seems to be doing it, I'm going to start posting a discussion thread for each of the NVIDIA Game Ready Driver releases, starting with NVIDIA Driver 556.12 today. Most of this is copied from the corresponding public post on Reddit.

Game Ready Driver 556.12

Download driver here

New Feature and Fixes in Driver 556.12

Game Ready - This new Game Ready Driver provides the best gaming experience for the latest new games supporting DLSS 3.5 technology including The First Descendant. Further support for new titles leveraging NVIDIA DLSS technology includes the latest update for PAYDAY 3 which supports DLSS 3.

Fixed Gaming Bugs

• [The Last of Us Part 1] Out of memory error with 555.xx drivers [4663766]
• [Halo Infinite] Crashing during initial loading screen with 555.99 driver [4685335]
• [Valorant] In-game statistics incorrectly shows high latency with 555.xx drivers [4668309]

Fixed General Bugs

• Notebook mux switch is blocked when Overwolf application is running [3804893]

Notes

• Security Bulletin issues listed by Nvidia in the 'NVIDIA GPU Display Driver - July 2024' bulletin are addressed by 556.12 and later - NVIDIA Security Bulletin
• 555.85 and later NVIDIA display drivers no longer support CPUs without the POPCNT (Population Count) instruction - NVIDIA Support
• HDR RTX multi-monitor requires both driver support and NVIDIA app support, the feature may be available with the next major version of the Nvidia App - NVIDIA Forum
• NVIDIA Broadcast Hotfix for Windows 11 24H2 is available at NVIDIA Broadcast Hotfix
• Chromium Checkboard issue, Microsoft have officially released a patch to the Windows 11 stable channel as KB5039212 (OS Builds 22621.3737 / 22631.3737) and Windows 10 stable channel as KB5040427 (OS Builds 19044.4651 and 19045.4651)
• Watch Dogs 2 flickering issue fix, see 'Troubleshooting and Workarounds' section
• Halo Wars 2 in-game foliage issue fix, see 'Troubleshooting and Workarounds' section
• NVGPUCOMP64.dll and/or 'out of VRAM' crashes, see 'Troubleshooting and Workarounds' section
• Nvidia App (BETA) may impact frame rates (?CPU bound scenarios?) on some systems, this isn't a Nvidia driver issue

Branch Information

Recent branch information only

• 560.38: r560_bugfix_main (Insider)
• 556.12: r556_09-02 (Game Ready)
• 555.99: r555_97-02 (Game Ready/Quadro/Studio)
• 555.85: r555_79-02 (Game Ready/Studio)
• 552.61: VK551_06-15 (Developer)
• 552.44: r552_19-08 (Game Ready)
• 552.22: r552_19-02 (Game Ready/Studio)
• 551.86: r550_00-192 (Game Ready/Studio)

Open Issues

• [NVENC] Quality and bitrate settings are ignored at 10-bit encoding [4697900]
• [GeForce Experience] Performance Monitoring overlay may stop refreshing GPU information [4679970]

Confirmed Additional Issues

This section covers issues officially acknowledged by Nvidia that are not in the original driver release notes:

• [GeForce RTX Series 40] graphics cards running older firmware could experience blank screens on boot with certain motherboards in UEFI mode until the OS loads - NVIDIA Support
• [HVEC/AV1 Playback] when playing back an HVEC/AV1 10-bit 4:2:0 file and scaling the viewport down to < 50%, video may appear slightly blurry [4496901]
• [GeForce Experience/NVIDIA App] if GPU utilization gets stuck at 0% inside of the GeForce Experience/NVIDIA App Performance overlay, games may display stutter [4720547] UPDATE: workaround uninstall GeForce Experience/NVIDIA app then restart system OR revert to driver 552.44 and older
• [Halo Infinite] crashing during initial loading screen on Windows 11 24H2 OS only [4685335] UPDATE: Fix/workaround in next Game Ready Driver - Reddit Discussion

Unconfirmed Additional Issues

This section contains issues not officially acknowledged by Nvidia but are reported across multiple forums:

• [GeForce RTX Series 40 GPUs] stability/TDR/black screen issues. Check for a motherboard BIOS update that states 'compatibility updates for Lovelace'. The motherboard update is in addition to any VBIOS update
• [Microsoft Store/GamePass/Xbox Application] PC games may not have Nvidia 'game ready driver profile' settings applied due to the APPID not being present in the Nvidia game profile

Looks like Valve is making progress towards releasing SteamOS on other platforms.

Original blog: https://zamundaaa.github.io/wayland/2024/04/05/explicit-sync.html

#What is Nvidia Blackwell? Named after David Blackwell, the first black scholar inducted into the National Academy of Science, Blackwell (GB200, which stands for Grace-Blackwell 200) follows the success of its predecessor, Hopper(GH100, which stands for Grace-Hopper 100), which boosted Nvidia’s sales.

For anybody who is not familiar with the naming conventions of Nvidia’s chip world, G or Grace is the CPU used in Nvidia’s AI computing modules called DGX, and B or Blackwell is a GPU, which is an upgrade from the previous H or Hopper.

Nvidia Blackwell and the whole DGXGB200 SuperPOD AI computer Nvidia Blackwell and the whole DGXGB200 SuperPOD AI computer Apart from the apparent benchmark scores that Nvidia proudly wears on its chest, the most assuming part is that Nvidia has taken two dies or chips and just merged them without any loss of computational capabilities, giving birth to GB200, which is insane from an engineering perspective.

One of the direct competitors, AMD MI300, will indeed be shivering when they hear these numbers.

The new Blackwell ship is 4–5 times faster than its predecessor, Hopper. But when integrated into its environment of DGX Module, it becomes a beast and outputs almost 30x the inference performance, yes, not 30% but 30 times, and wait, wait, there is more, all while consuming 25X less power again that is not 25% its 25 times. This allows LLM’s models to train on over 27 trillion parameters. For comparison, the famous ChatGPT 4.0 was trained on 1.7 trillion parameters. Blackwell will be made in partnership with Nvidia’s best friend, TSMC, during a fantastic 3 nm fabrication process. It has 208 billion transistors, allowing data speeds up to 10Tb/sec. Amazing right?

#How is Blackwell(GB200) better than Hopper GPU(GH100)?

The previous generation of AI-optimized GPUs was called Hopper. Blackwell is between 2 and 30 times faster, depending on how you measure it. Huang explained that it took 8,000 GPUs, 15 megawatts, and 90 days to create the GPT-MoE-1.8T model. With the new system, you could use just 2,000 GPUs and 25% of the power.

Blackwell is genuinely a very successful replacement for the Hopper GPU because it builds upon GH100’s strengths and fills in the gaps or limitations, like improved communication bottlenecks for a larger number of racks. It has no Memory locality issues or cache storage issues.

This essentially means there are much lower bottlenecks when sharing data between servers, dramatically reducing the idle time for each GPU on each server. Thus, efficiency is boosted since more stuff gets done in the same clock cycles.

GB200 offers improved connectivity and data processing for AI tasks and is part of Nvidia’s “super chip” lineup, complementing its central processing unit, Grace.

Nvidia will definetly help AI to reach new heights. Click the link to visit and read the full blog.