Large language models and the applications they power enable unprecedented opportunities for organizations to get deeper insights from their data reservoirs and to build entirely new classes of applications.
But with opportunities often come challenges.
Both on premises and in the cloud, applications that are expected to run in real time place significant demands on data center infrastructure to simultaneously deliver high throughput and low latency with one platform investment.
To drive continuous performance improvements and improve the return on infrastructure investments, NVIDIA regularly optimizes the state-of-the-art community models, including Meta's Llama, Google's Gemma, Microsoft's Phi and our own NVLM-D-72B, released just a few weeks ago.
Relentless Improvements Performance improvements let our customers and partners serve more complex models and reduce the needed infrastructure to host them. NVIDIA optimizes performance at every layer of the technology stack, including TensorRT-LLM, a purpose-built library to deliver state-of-the-art performance on the latest LLMs. With improvements to the open-source Llama 70B model, which delivers very high accuracy, we've already improved minimum latency performance by 3.5x in less than a year.
We're constantly improving our platform performance and regularly publish performance updates. Each week, improvements to NVIDIA software libraries are published, allowing customers to get more from the very same GPUs. For example, in just a few months' time, we've improved our low-latency Llama 70B performance by 3.5x.
NVIDIA has increased performance on the Llama 70B model by 3.5x. In the most recent round of MLPerf Inference 4.1, we made our first-ever submission with the Blackwell platform. It delivered 4x more performance than the previous generation.
This submission was also the first-ever MLPerf submission to use FP4 precision. Narrower precision formats, like FP4, reduces memory footprint and memory traffic, and also boost computational throughput. The process takes advantage of Blackwell's second-generation Transformer Engine, and with advanced quantization techniques that are part of TensorRT Model Optimizer, the Blackwell submission met the strict accuracy targets of the MLPerf benchmark.
Blackwell B200 delivers up to 4x more performance versus previous generation on MLPerf Inference v4.1's Llama 2 70B workload. Improvements in Blackwell haven't stopped the continued acceleration of Hopper. In the last year, Hopper performance has increased 3.4x in MLPerf on H100 thanks to regular software advancements. This means that NVIDIA's peak performance today, on Blackwell, is 10x faster than it was just one year ago on Hopper.
These results track progress on the MLPerf Inference Llama 2 70B Offline scenario over the past year. Our ongoing work is incorporated into TensorRT-LLM, a purpose-built library to accelerate LLMs that contain state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT-LLM is built on top of the TensorRT Deep Learning Inference library and leverages much of TensorRT's deep learning optimizations with additional LLM-specific improvements.
Improving Llama in Leaps and Bounds More recently, we've continued optimizing variants of Meta's Llama models, including versions 3.1 and 3.2 as well as model sizes 70B and the biggest model, 405B. These optimizations include custom quantization recipes, as well as efficient use of parallelization techniques to more efficiently split the model across multiple GPUs, leveraging NVIDIA NVLink and NVSwitch interconnect technologies. Cutting-edge LLMs like Llama 3.1 405B are very demanding and require the combined performance of multiple state-of-the-art GPUs for fast responses.
Parallelism techniques require a hardware platform with a robust GPU-to-GPU interconnect fabric to get maximum performance and avoid communication bottlenecks. Each NVIDIA H200 Tensor Core GPU features fourth-generation NVLink, which provides a whopping 900GB/s of GPU-to-GPU bandwidth. Every eight-GPU HGX H200 platform also ships with four NVLink Switches, enabling every H200 GPU to communicate with any other H200 GPU at 900GB/s, simultaneously.
Many LLM deployments use parallelism over choosing to keep the workload on a single GPU, which can have compute bottlenecks. LLMs seek to balance low latency and high throughput, with the optimal parallelization technique depending on application requirements.
For instance, if lowest latency is the priority, tensor parallelism is critical, as the combined compute performance of multiple GPUs can be used to serve tokens to users more quickly. However, for use cases where peak throughput across all users is prioritized, pipeline parallelism can efficiently boost overall server throughput.
The table below shows that tensor parallelism can deliver over 5x more throughput in minimum latency scenarios, whereas pipeline parallelism brings 50% more performance for maximum throughput use cases.
For production deployments that seek to maximize throughput within a given latency budget, a platform needs to provide the ability to effectively combine both techniques like in TensorRT-LLM.
Read the technical blog on boosting Llama 3.1 405B throughput to learn more about these techniques.
Different scenarios have different requirements, and parallelism techniques bring optimal performance for each of these scenarios. The Virtuous Cycle Over the lifecycle of our architectures, we deliver significant performance gains from ongoing software tuning and optimization. These improvements translate into additional value for customers who train and deploy on our platforms. They're able to create more capable models and applications and deploy their existing models using less infrastructure, enhancing th
Most recent headlines
05/01/2027
Worlds first 802.15.4ab-UWB chip verified by Calterah and Rohde & Schwarz to be ...
06/09/2026
June 9 2026, 23:00 (PDT) Dolby and MagentaTV Bring Fans Closer to the FIFA Worl...
04/08/2026
Dalet, a leading technology and service provider for media-rich organizations, t...
06/07/2026
Long-developed relationships and informal access, three production models, and c...
06/07/2026
The deal follows Lumine's recent acquisition of Synamedia's video-networ...
06/07/2026
This former SportsCenter producer is shaping local-team studio coverage to surro...
06/07/2026
IMS Productions and FOX Sports deployed more than 100 cameras, 32 onboard feeds,...
06/07/2026
Lumine Group Inc. has announced the completion of its acquisition of the Video Network Business from Synamedia. The business will operate as a standalone compan...
06/07/2026
NASCAR has announced a content and commercial partnership with Motorsport Network, a motorsports and automotive media platform, to distribute NASCAR original co...
06/07/2026
Versant Media Group has announced a definitive agreement to acquire Full Swing, a sports technology company, from Bruin Capital and a group of minority investor...
06/07/2026
DAZN has announced global broadcast rights to 3ICE, a professional hockey format featuring three skaters and a goalkeeper per side, with no power plays or offsi...
06/07/2026
Bundle gains 100 new synths & more
Synth Anthology 5 the largest and most feature-packed edition of UVI's hardware synth emulation bundle, bringing toge...
06/07/2026
Available now via free firmware update
MOTU have just announced that three of their popular audio interfaces now boast official Milan Certification. The 16A...
06/07/2026
Rohde & Schwarz completes Hunter Preliminary Design Review for the Integrated Co...
06/07/2026
The R&S IMAGER from Rohde & Schwarz uses millimeter wave imaging to inspect the ...
06/07/2026
The National Film and Video Foundation (NFVF), an agency of the Department of Sp...
06/07/2026
Digital Domain Appoints Industry Veteran Patrick Davenport as Executive Vice Pre...
06/07/2026
Gravity Media USA appoints Brittney Boston as Head of Business Development
Brie Clayton July 6, 2026
0 Comments
The strategic new appointment supports...
06/07/2026
Remote Editing Without Cloud Storage in Resolve
Michael Cioni July 6, 2026
0 Comments
This is a step-by-step guide that demonstrates how to setup your...
06/07/2026
Automatic Caption Translation in Adobe Premiere Pro
Colin Smith July 6, 2026
0 Comments
Here I showcase the innovative automatic caption translation f...
06/07/2026
Flavor's First Ever Managing Director Neal Cohen Retires; Longstanding Execu...
06/07/2026
How Adam Rosenwach Pivoted from Music to Medtech Without Missing a Beat What do the rehearsal room and the boardroom have in common? More than you might think...
06/07/2026
Zhongyi Optics unveils the Dreamy Zone T1 FF Cine Lens Series on Indiegogo with ...
06/07/2026
Techtel debuts insyght: making ST 2110 media discovery easy
intellisphereinsyght2026
6 July Written By Suzanne Costello
Sydney, Australia - 06 July 2026 T...
06/07/2026
Monday 6 July 2026
Sky agrees to acquire ITV Media & Entertainment, creating a ...
06/07/2026
Fresh from his World Cup heroics with Cabo Verde, Roberto Pico' Lopes will join the RT panel for Argentina v Egypt tomorrow and for the England v Norway q...
06/07/2026
Nations have long invested in domestic infrastructure to advance their economies, protect and use their data, and take advantage of technology opportunities in ...
06/07/2026
Every year, the International Conference on Machine Learning (ICML) reveals where thousands of AI researchers have decided to put their work.
This year's ...
06/07/2026
A death can be unsuspicious, but it has to be explained
What happened to Ke...
05/07/2026
We've gone to another level of integrating the graphics into the set architecture' - Zac Fields, SVP, Graphic Technology and Integration, FOX Sports...
05/07/2026
Browser-based and desktop apps available
Digital Sunset Studios have been busy working on an interesting project: a free software tool that provides compute...
04/07/2026
When Canada qualified for the FIFA World Cup 2026, OneSoccer knew this wasn'...
04/07/2026
Celebrates company's 80th anniversary
Rhodes have recently revealed that they will be producing a limited run of electric pianos in celebration of their...
04/07/2026
Blackmagic Design Cameras Empower Youth Broadcasting Program
Brie Clayton July 3, 2026
0 Comments
Blackmagic Pocket Cinema Camera 6K Pro and Blackmagi...
04/07/2026
iZotope Joins Boris FX
Boris Yamnitsky July 3, 2026
0 Comments
iZotope, the team behind RX and Ozone, is joining Boris FX. A letter from founder Boris...
04/07/2026
April 7 2026, 19:00 (PDT) Detective Conan: Fallen Angel of the Highway Opens in...
03/07/2026
New bundle & three Single Packs
Continuing to expand their already sizeable orchestral collection, VSL's latest release introduces three new Single Pack...
03/07/2026
Venue installs new ATC-equipped control room
LSO St Luke's, a unique music venue that also serves as the home of the London Symphony Orchestra, have rec...
03/07/2026
Caden Pearson appointed new Commissioning Editor for NITV
3 July, 2026
Media releases
National Indigenous Television (NITV) has strengthened its commitment...
03/07/2026
1 February 2023
SHARE Facebook Twitter Linkedin Email
Munich, Germany, 1st February 2023: Cinegy GmbH, the premier provider of software technology for digit...
03/07/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
03/07/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
03/07/2026
Share
Copy link
Facebook
X
Linkedin
Bluesky
Email...
03/07/2026
Death Star
Andy Marken July 2, 2026
0 Comments
Look, I can't get involved. I've got work to do. It's not that I like the Empire; I hate i...
03/07/2026
Berklee's New Visual Identity: Honoring Our History, Building for What's...
03/07/2026
Scripps Research scientists awarded $2M to advance global disease surveillance Two Gates Foundation grants will expand wastewater surveillance and AI-driven dis...
03/07/2026
Joan Pulupa joins Scripps Research faculty to study the organization of DNA in brain cells and its links to neurodegeneration Using smell-sensing neurons and ad...
02/07/2026
Entering her senior year, this hometown girl is paving a career in live sports production gaining experience in replay and audio and as a TD
In the live-sports...
02/07/2026
In-venue and creative video staffers at the professional and collegiate level ha...
Large language models and the applications they power enable unprecedented opportunities for organizations to get deeper insights from their data reservoirs and to build entirely new classes of applications.
