Join us on the road to #exascale El Capitan! Our multi-part article series continues with installments 6 through 10. 🎉

computing.llnl.gov/livermore-comp…

#HPC #supercomputer #computing

AMD and GigaIO invite you to shatter the 8-GPU server ceiling without changing a single line of code.

Learn how can you shorten time-to-results when training large models, while saving power and space in your #datacenter .

Join us on April 10 at 9am PT.
bit.ly/49foKei

This is an elite team, pushing the limits of scalable computing in AI.

Andres Freund, the principal software engineer at Microsoft who discovered the xz backdoor really does deserve a big pat on the back. 👏

The outcome could have been much, much worse.

It will be my great pleasure to present this award at ISC HPC. This is a great paper on a great topic.

LLNL distinguished member of technical staff & Spack lead Todd Gamblin / @[email protected] is one of the HPCwire #PeopletoWatch 😎

llnl.gov/article/51011/…

#HPC #computing #nationallab #opensource

Well...... I am quite underwhelmed by GB200.....
Nvidia cut FP64 massively and it's quite depressing....
A single Blackwell GPU only gets about 20-22.5TF of FP64 Tensor (And no I am not counting Sparsity numbers here) which means that the Vector FP64 is only 10-11TF......

Looks like FP64 was borderline culled on B100

using the attached table, that's 90 SPARSE Tflops of FP64 using 2 GPUs, so 'peak' of ~22.5 Tflops per device

lop the typical GPGPU Max/Peak ratio on and we're probably looking towards ~15Tflops of dense, double precision MatMul
#HPC

Great time chatting with George Markomanolis

Bruce Schneier on 'How Public AI can Strengthen Democracy' -- 'Federally funded foundation AI models would be provided as a public service, similar to a health care private option. They would not eliminate opportunities for private foundation models, but they would offer a

Why America Must Invest in DOE Labs To Win the #AI Race Against China

realcleardefense.com/articles/2024/…

#HPC U.S. Department of Energy DOE Office of Science

A little update: I will be joining AMD for this summer as a PhD tech intern, here in Austin. Very glad I won't be relocating and and it seems that I will be on a fantastic team!

name cannot be blank 🐀 HPC Guru (Taking a break) NVIDIA Nvidia as 100% of inference is just not true.

Mi250X, MI300, CPU by Intel/AMD and TPUs from Google and others.

AMD Hires Former Oak Ridge Director Thomas Zacharia
wp.me/p3RLHQ-olK
AMD AMDServer Oak Ridge Lab #HPC

We are hiring an HPC (MPI / OpenAI Triton) Engineer at Lamini. Apply here:

jobs.lever.co/laminiai/af688…

We are inventing and building the largest AMD LLM training system in the world. Join us in strongly scaling to 1000s of GPUs and beyond.