Missed @hpcnotes presentation at the @hpccouncil UK Conference where he covered #HPC @Azure and thoughts as we approach #Exascale?

You can watch it here:
Andrew makes some excellent points about the impact of doing things at extreme scale - whether it be #cloud or #HPC

Even a small number multiplied by a very large number results in something significant. Hence the need to optimize all aspects - he gives a good example - PUE.
Even if the large public #cloud providers may not have breached the #Exascale barrier (for those who follow me know what that is defined as 😎) , they are, as @hpcnotes said, "unquestionably at a large enough scale to share relevant lessons for Exascale"

#HPC
What issues do hyperscale #cloud or #Exascale #HPC bring?

What is the cost of getting it right vs. the cost/risk of getting it wrong?

Scale matters!

Slide from @hpcnotes presentation
Perhaps the most interesting part of Andrew's presentation is comparing what was predicted for #Exascale vs. the reality.

In reality, it is much more of an evolution rather than a revolution!

The dominant programming model is still #MPI + X

#HPC
Things that we (the #HPC community) did not foresee in early predictions about #Exascale

o The emergence of #AI as a dominant use case

o Vendor consolidation (HPE buys both SGI and Cray, IBM seems to have lost interest)

o Diversity in processors

/End of thread
You can follow @HPC_Guru.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled:

By continuing to use the site, you are consenting to the use of cookies as explained in our Cookie Policy to improve your experience.