Percentage wirelength reduction of the RL solution vs. the same block placed by a human expert.
HARDWARE
-
Silent Data Corruption Risks in Large-Scale ML Training
By
–
I discussed the challenges of silent data corruption in ML training jobs, and how one faulty piece of hardware can infiltrate and affect the results of a large scale training jobs on thousands of chips.
-
Reinforcement Learning Advances TPU Chip Design Across Generations
By
–
Among many other topics, here are a few things from the talk. I gave an update on our use of reinforcement learning for ASIC chip placement and routing, showing how the adoption of this approach has been increasing from TPU generation to generation (due to further work on the
-
SPUs: Soul-Powered Computing 10^10x Faster Than H100s
By
–
SPUs (Soul Processing Units) are 10^10 times more powerful than H100s but takes a human soul to operate for a full day
-
NVIDIA leads Japan’s generative AI infrastructure development initiative
By
–
Following an announcement by Japan’s Ministry of Economy, Trade and Industry, NVIDIA will play a central role in developing the nation’s generative AI infrastructure as Japan seeks to capitalize on the technology’s economic potential and further develop its workforce. Learn
-
Cutting-Edge AI Video Analysis Hardware Solutions Showcase
By
–
We had a blast showcasing our cutting-edge AI video analysis solutions & connecting with the HPC community at #ISC24. Missed us? Catch us next week at @edgeaivision
's #EVS24, where we'll demo how our hardware is revolutionizing AI performance across industries! -
Cerebras Accelerates Molecular Dynamics 179X Faster Than Frontier
By
–
@TDaytonPM of @TheNextPlatform writes how researchers from Cerebras and U.S. National Labs have accelerated a molecular dynamics application by a factor of 179X compared to the same application running on the 'Frontier' supercomputer at Oak Ridge National Laboratory. "On the
-
179x Speedup Breakthrough in Molecular Dynamics with CS-2
By
–
🎉 Our MLSys Mixer was a smash!🎉
— Cerebras (@cerebras) 16 mai 2024
We had over 100+ register and join on the eve of our monumental announcements!
In case you missed it, we announced the following:
📢 Molecular Dynamics Breakthrough: 179x speedup over Frontier Supercomputer using a single CS-2 -… pic.twitter.com/kMHDhNyL34Our MLSys Mixer was a smash! We had over 100+ register and join on the eve of our monumental announcements! In case you missed it, we announced the following: Molecular Dynamics Breakthrough: 179x speedup over Frontier Supercomputer using a single CS-2 –
-
BioHive AI Engine Advances 100 Spots TOP500 Supercomputers
By
–
BioHive, an engine powered by NVIDIA #AI to speed the work of scientists in healthcare, rose more than a hundred spots on the TOP500 list of the world’s fastest supercomputers. Learn more: https://
nvda.ws/44KFLwa -
NVIDIA Tools Scale Generative AI and LLM Deployment
By
–
Deploying generative #AI at scale is complex, but NVIDIA provides the essential tools, frameworks, and models to quickly take LLMs and generative AI applications to market. Join our session at #MSBuild to get started. Register today > https://
nvda.ws/3wyjK7i