Join Nicholas Malaya of AMD & @john_ihnotic from #GigaIO to learn to speed up model training by breaking the 8-GPU server limit. No code changes needed. Join us April 10, 9am PT for insights on #nextgenAI & #acceleratedcomputing . bit.ly/3xhReqz #ServerCeiling

The first to deploy MI300X publicly at scale. 🚀

#amd #mi300x #gpu #gpu cloud #rocm

I asked Lenovo's North American president about demand for AMD's Instinct MI300X GPU, and he gave a pretty interesting response. Note that this was before Nvidia's Blackwell announcement, but I think the comment is still very relevant. crn.com/news/computing… #amd #nvidia

Great Tech Blog from Dell on 'Silicon Diversity: Deploy GenAI on the PowerEdge XE9680 with AMD Instinct MI300X Accelerators' | Dell Technologies Info Hub infohub.delltechnologies.com/en-US/p/silico…

Do you remember when you joined X? I do! #MyXAnniversary

And it’s running on AMD Instinct MI300

A technical deep dive into how we set up multi-node training on AMD GPUs and speed up LLM training for 1000x or even 10,000x! Led by our amazing ayushi sharma

👉 lamini.ai/blog/multi-nod…

Great to see Phoronix get some hands-on time with AMD Instinct MI300X and ROCm.

35 seconds of CFD simulation were ran on OLCF #Frontier AMD EPYC 64c CPUs and AMD Instinct MI250X accelerators. There was a feedback loop implemented to use the simulated data to run a simulated control system. Fantastic visualizations there!

🎥 : OLCF Director of Science Bronson Messer joined WATE 6 On Your Side to discuss the Frontier supercomputer's role in landing humans safely on Mars. Oak Ridge Lab has been working with NASA to support this mission since 2019. Bronson Messer bit.ly/3IlSmM9

🚨New Website Alert🚨

Communications of the ACM @cacmmag has relaunched its website! First published in 1958, our flagship magazine is transitioning into a fully #OpenAccess , web-first publication.

Check it out NOW🚀: bit.ly/3Dj5JKc

According to StatCounter's data, by the end of February 2024, #Linux has achieved a 4.03% desktop market share.
linuxiac.com/linux-crosses-…

👇Lamini says they are building the largest AMD #LLM training system in the world

In a market dominated by Nvidia, credible competition is badly needed

#HPC #AI #GenAI AMDInstinct via Greg Diamos

The Italian Energy company ENI has announced a major #HPC system based on AMD Instinct MI250X GPUs, EPYC CPUs and will be an HPE Cray EX4000 (like Frontier, LUMI, etc.)

eni.com/en-IT/media/pr…

Thanks to Essam Morsi and the AMD team for a very exciting project!

This is the largest LLM training example I'm aware of on AMD AMDInstinct hardware. And likely some of the largest LLM training in the world.

Further proof of the important impact HPC and Supercomputing is having on driving new capabilities in science and AI.

This is 'just' 3k of the 37k MI250X on Frontier. To steal from Feynman, 'There is plenty of room at the top!' -- I'm expecting lots of interesting work scaling out further to tens of thousands of nodes.

Really exciting results in training AI models on Frontier, the world's largest supercomputer. This is a 1 Trillion parameter LLM run on 3k Frontier MI250X GPUs. To put this in perspective: 1 trillion parameters is on the scale of chatGPT/GPT-4.

Paper at: lnkd.in/gX7rExXp

On this day in 1939, HP launched in this Palo Alto garage. It is considered the birthplace of Silicon Valley.

Great list to save.