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...
09/06/2026
Media buyers and sellers can now compare YouTube reach from computer, mobile, an...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
A project aimed at modernizing scheduling operations across Mediaset's channel portfolio
Mediagenix, a global leader in smart content solutions to profitab...
09/06/2026
tvONE, an ACT Entertainment brand, debuts its most powerful video processor ever built: the 4RU CALICO PRO (C7-PRO-4200), at InfoComm 2026 (Booth N6813). Engine...
09/06/2026
Composer and Conductor Eric Whitacre to Receive Honorary Doctorate at Berklee Va...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
Technology collaboration to validate AMPP and Lawo HOME orchestration integration, showcasing Dynamic Media Facility principles in practice.
Grass Valley and ...
09/06/2026
Bitmovin, a leading provider of video streaming solutions, today announced that Axel Springer SE, international media and technology company, has deployed Playe...
09/06/2026
LiveU, the global leader in live IP-video solutions, today announced a landmark deployment with the Oceania Football Confederation (OFC) that has brought Video ...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
09/06/2026
Two Alumni Win Tony Awards Mike Morris and Cedric Leiba Jr. won awards for Best Orchestrations and Best Play, respectively.
June 8, 2026
By
Tori Donahue
...
09/06/2026
Enni Continues to Rely on Arvato Systems - And Will Be Using AEP.EnerS4 Billing Solution
Arvato Systems to implement transformation project for SAP S/4HANA Ut...
09/06/2026
These special programmes will highlight RT 's connection with the Irish diaspora
This year, RT is marking 100 years of public broadcasting in Ireland with...
08/06/2026
The two execs discuss the goals and the challenges of building a huge multi-cult...
08/06/2026
This industry legend has been a road warrior since the early '90s and is, today, a lynchpin on College GameDay
The sports-production industry continues to ...
08/06/2026
DAZN has agreed a multi-year rights partnership with Extreme Fighting Championship (EFC), expanding its global combat sports portfolio.
DAZN will broadcast EFC...
08/06/2026
As fans move across broadcast, streaming, social, mobile and creator-led platforms, and they expect every angle, in real time, TVU Networks has released a new e...
08/06/2026
Infront has further strengthened the 2026 Diamond League's international bro...
08/06/2026
Grass Valley has announced the appointment of Sam Craig as Vice President, Global Pre-Sales. Based in the United States, Craig will lead Grass Valley's glob...
08/06/2026
SMPTE has announced its summer 2026 education lineup covering IP networking and SMPTE ST 2110, including two standalone courses, a 13-week boot camp, and a hand...
08/06/2026
Sportradar Group AG has announced a multi-year global agreement with Kalshi, a p...
08/06/2026
FIFA and Gamefam have announced the launch of a FIFA World Cup 2026 event on Roblox, running June 5 through July 31, 2026. The event centers on FIFA Super Socce...
08/06/2026
A collaboration of NHL Productions, Monumental Sports, and Michael Levitt Produc...
08/06/2026
CBS Sports production staffer Bryce Adair has died at the age of 31 after sustaining injuries during a car accident on Wednesday while in Ohio working the the P...
08/06/2026
Behind The Mic provides a roundup of recent news regarding on-air talent, including new deals, departures, and assignments compiled from press releases and repo...
08/06/2026
WMT Digital today announced the launch of WMT Fan Intelligence, designed to unif...
08/06/2026
The Heat are the seventh NBA team to air their games over-the-air next season...
08/06/2026
Paramount and UFC have announced an expansion of their media rights partnership making Paramount the exclusive home of UFC Numbered Event main cards in Canada ...
08/06/2026
The Qatar Football Association (QFA), in cooperation with Sponix and broadcaster...
08/06/2026
Project B has announced the appointment of Host Broadcast Services (HBS) as Offi...
08/06/2026
The Buckeyes now produce and stream upwards of 200 sports events annually with a...
08/06/2026
At Spotify, our commitment to the LGBTQIA community is year-round. Through GLOW, our global music program, we celebrate and amplify the contributions of queer ...
08/06/2026
Get Hands-on With Keyboard & Synth Brands
GearExpo UK wouldn't be complete without some synth action, and we've got some of the industry's most ...
08/06/2026
50 popular in-ear monitoring system profiles added
The latest update for IK Multimedia's headphone-correction system has just arrived, and introduces ca...
08/06/2026
Manny Marroquin signature cans upgraded with SLAM Technology
The flagship model in Audeze's Manny Marroquin Signature Series has just been treated to an...
08/06/2026
Air domain supremacy redefined - New counter UAS, space solutions and directiona...
08/06/2026
he BBC and BritBox have announced that Edward Bluemel (We Might Regret This, My ...
08/06/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
08/06/2026
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.
