The full-stack NVIDIA accelerated computing platform has once again demonstrated exceptional performance in the latest MLPerf Training v4.0 benchmarks.
NVIDIA more than tripled the performance on the large language model (LLM) benchmark, based on GPT-3 175B, compared to the record-setting NVIDIA submission made last year. Using an AI supercomputer featuring 11,616 NVIDIA H100 Tensor Core GPUs connected with NVIDIA Quantum-2 InfiniBand networking, NVIDIA achieved this remarkable feat through larger scale - more than triple that of the 3,584 H100 GPU submission a year ago - and extensive full-stack engineering.
Thanks to the scalability of the NVIDIA AI platform, Eos can now train massive AI models like GPT-3 175B even faster, and this great AI performance translates into significant business opportunities. For example, in NVIDIA's recent earnings call, we described how LLM service providers can turn a single dollar invested into seven dollars in just four years running the Llama 3 70B model on NVIDIA HGX H200 servers. This return assumes an LLM service provider serving Llama 3 70B at $0.60/M tokens, with an HGX H200 server throughput of 24,000 tokens/second.
NVIDIA H200 GPU Supercharges Generative AI and HPC The NVIDIA H200 Tensor GPU builds upon the strength of the Hopper architecture, with 141GB of HBM3 memory and over 40% more memory bandwidth compared to the H100 GPU. Pushing the boundaries of what's possible in AI training, the NVIDIA H200 Tensor Core GPU extended the H100's performance by up to 47% in its MLPerf Training debut.
NVIDIA Software Drives Unmatched Performance Gains Additionally, our submissions using a 512 H100 GPU configuration are now up to 27% faster compared to just one year ago due to numerous optimizations to the NVIDIA software stack. This improvement highlights how continuous software enhancements can significantly boost performance, even with the same hardware.
This work also delivered nearly perfect scaling. As the number of GPUs increased by 3.2x - going from 3,584 H100 GPUs last year to 11,616 H100 GPUs with this submission - so did the delivered performance.
Learn more about these optimizations on the NVIDIA Technical Blog.
Excelling at LLM Fine-Tuning As enterprises seek to customize pretrained large language models, LLM fine-tuning is becoming a key industry workload. MLPerf introduced a new LLM fine-tuning benchmark this round, based on the popular low-rank adaptation (LoRA) technique applied to Meta Llama 2 70B.
The NVIDIA platform excelled at this task, scaling from eight to 1,024 GPUs, with the largest-scale NVIDIA submission completing the benchmark in a record 1.5 minutes.
Accelerating Stable Diffusion and GNN Training NVIDIA also accelerated Stable Diffusion v2 training performance by up to 80% at the same system scales submitted last round. These advances reflect numerous enhancements to the NVIDIA software stack, showcasing how software and hardware improvements go hand-in-hand to deliver top-tier performance.
On the new graph neural network (GNN) test based on R-GAT, the NVIDIA platform with H100 GPUs excelled at both small and large scales. The H200 delivered a 47% boost on single-node GNN training compared to the H100. This showcases the powerful performance and high efficiency of NVIDIA GPUs, which make them ideal for a wide range of AI applications.
Broad Ecosystem Support Reflecting the breadth of the NVIDIA AI ecosystem, 10 NVIDIA partners submitted results, including ASUS, Dell Technologies, Fujitsu, GIGABYTE, Hewlett Packard Enterprise, Lenovo, Oracle, Quanta Cloud Technology, Supermicro and Sustainable Metal Cloud. This broad participation, and their own impressive benchmark results, underscores the widespread adoption and trust in NVIDIA's AI platform across the industry.
MLCommons' ongoing work to bring benchmarking best practices to AI computing is vital. By enabling peer-reviewed comparisons of AI and HPC platforms, and keeping pace with the rapid changes that characterize AI computing, MLCommons provides companies everywhere with crucial data that can help guide important purchasing decisions.
And with the NVIDIA Blackwell platform, next-level AI performance on trillion-parameter generative AI models for both training and inference is coming soon.
Most recent headlines
05/01/2027
Worlds first 802.15.4ab-UWB chip verified by Calterah and Rohde & Schwarz to be ...
04/08/2026
Dalet, a leading technology and service provider for media-rich organizations, t...
04/07/2026
April 7 2026, 19:00 (PDT) Detective Conan: Fallen Angel of the Highway Opens in...
01/06/2026
January 6 2026, 05:30 (PST) Dolby Sets the New Standard for Premium Entertainment at CES 2026
Throughout the week, Dolby brings to life the latest innovatio...
02/05/2026
Dalet, a leading technology and service provider for media-rich organizations, t...
01/05/2026
January 5 2026, 18:30 (PST) NBCUniversal's Peacock to Be First Streamer to ...
21/04/2026
Cloud-based production isnt going anywhere, and BitFire is doubling down by prov...
21/04/2026
The topic of artificial intelligence has a stranglehold on the sports-video-prod...
21/04/2026
5G is still a hot topic in live event production, and this workflow continues to...
21/04/2026
At the 2026 NAB Show, Ed McGivern, GM and President of Appear US, discusses the ...
21/04/2026
Studio Network Solutions (SNS) has announced an on-premise AI suite designed for...
21/04/2026
Suite Studios has integrated its file-streaming technology into the newly announced Frame.io Drive, a desktop application from Adobe company Frame.io. The colla...
21/04/2026
Net Insight has integrated InSync Technology's FrameFormer into the Nimbra E...
21/04/2026
Fox Sports has selected Appear as a technology partner to support the next phase...
21/04/2026
Diversified has appointed Tyler Affolter as Chief Revenue Officer (CRO) to lead the company's commercial organisation. The appointment follows the firm'...
21/04/2026
Layercake has formalised the integration of Bitmovin's video streaming infra...
21/04/2026
The International Judo Federation (IJF) has extended its distribution partnershi...
21/04/2026
Glookast has launched the Cinnafilm Tachyon plugin for its Media Producer and Me...
21/04/2026
Eutelsat has entered into an agreement with Cadena Tres, a division of Grupo Ima...
21/04/2026
Dolby Laboratories and TV Azteca have partnered to introduce Dolby Atmos immersive audio to free-to-air television broadcasts. The implementation utilises the A...
21/04/2026
FOX Entertainment partnered with Verizon to overcome significant production hurd...
21/04/2026
Osprey Video has announced its technology showcase for the NAB Show 2026, highli...
21/04/2026
Riedel Communications (Booth C4908) introduced a range of new solutions at NAB S...
21/04/2026
The NAB Show is in full swing, and the SVG and SVG Europe editorial teams are chasing down the hottest stories from all over the Las Vegas Convention Center. He...
21/04/2026
Blackmagic Design has announced the URSA Cine Immersive 100G, an immersive cinem...
21/04/2026
Clark Wire & Cable is continuing its evolution from cable supplier to full-scale solutions partner for broadcast and live production. At the 2026 NAB Show, we s...
21/04/2026
Rashad Frett attends the 2025 Sundance Film Festival premiere of Ricky at Eccles Theatre on January 24, 2025, in Park City, UT. (Photo by George Pimentel/Shut...
21/04/2026
Five years ago, Spotify arrived in Pakistan, opening a new chapter in the country's music scene. Since then, local listeners have explored across genres, ge...
21/04/2026
Since its launch in 2020, RADAR has been our program for spotlighting emerging artists around the world. This year marks the sixth edition of RADAR Spain, our o...
21/04/2026
Offers compact wireless solution for pedalboards
Taiwanese audio brand Cloudvocal have announced the availability of a new pedalboard-friendly wireless syst...
21/04/2026
Latest hybrid sampling/synthesis instrument arrives
Arturia's Augmented series offerings rely on a mixture of sampling and synthesis, allowing users to ...
21/04/2026
Rohde & Schwarz to host Power Electronics Online Conference From Design to Vali...
21/04/2026
MAS and Lockheed Martin partner to establish an F-35 depot in Canada, enabling in-country sustainment and creating high-skilled aerospace jobs....
21/04/2026
Advertising strategies shift as competition grows for a large, active and qualit...
21/04/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
21/04/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
21/04/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
21/04/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
21/04/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
21/04/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
21/04/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
21/04/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
21/04/2026
Layercake Deepens Bitmovin Integration to Power End-to-End Media Orchestration w...
21/04/2026
Cobalt Digital Announces Launch of blueCORE at NAB Show 2026: Compact Powerhouse...
21/04/2026
On-Premise AI Suite from Studio Network Solutions Debuts at NAB Show 2026
Melanie Ciotti April 21, 2026
0 Comments
Unlimited processing, no cloud depe...
21/04/2026
London, 21 April 2026 IBC today announced the appointment of Tim Banham as its first Chief Commercial Officer (CCO), a newly created role that reflects the or...
21/04/2026
Motion Design Tools - April 2026
Roland Kahlenberg April 21, 2026
0 Comments
Within 2 days, Maxon and Canva announced pro-level motion design apps - A...
21/04/2026
Chaos and Zero Density to Showcase Real-Time Ray Tracing for Virtual Studios and...
21/04/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
The full-stack NVIDIA accelerated computing platform has once again demonstrated exceptional performance in the latest MLPerf Training v4.0 benchmarks.
