Andrew Jones (@hpcnotes) 's Twitter Profile
Andrew Jones

@hpcnotes

25 yrs using/researching/buying #supercomputers. Now an engineering leader for #supercomputing capabilities at Microsoft. Posts on #HPC #AI #cloud #F1 #travel

ID: 111317745

linkhttps://bsky.app/profile/hpcnotes.bsky.social calendar_today04-02-2010 14:17:35

17,17K Tweet

7,7K Takipçi

1,1K Takip Edilen

Daniel Newman (@danielnewmanuv) 's Twitter Profile Photo

$AMD can see its GPU numbers rise 8-10x to at least $40 billion a year by 2030 without taking any market share from $NVDA The idea that NVIDIA has to lose for AMD to win is incorrect. The market will allow multiple very large winners, but we don't see NVIDIA ceding considerable

Andrew Jones (@hpcnotes) 's Twitter Profile Photo

You know when you watched a recording of a meeting back at 2x speed? Anyone else feel like their whole week is being played at 2x speed and many days/meetings are just racing past?

Andrew Jones (@hpcnotes) 's Twitter Profile Photo

Procrastinating thought about #supercomputing … When does the world of #HPC lose the most productivity temporarily - during November when everyone is focusing on #SC25 - or during July/August when large numbers of people are taking holidays?

Andrew Jones (@hpcnotes) 's Twitter Profile Photo

Thought experiment: You have a #supercomputer with infinite FLOPS (FP64 or whatever else you prefer) and infinite memory bandwidth - what is now the biggest bottleneck on your science advances? (Or business value generation, or whatever your supercomputing mission is) #HPC #AI

Andrew Jones (@hpcnotes) 's Twitter Profile Photo

A question for chip gurus to explain (ie argue 😁) … Purely in architectural terms (ie ignore current success / adoption etc) - which is/was the best architecture: #x86 vs #ARM vs #RISCV vs #MIPS vs #SPARC vs #IA64 vs Alpha - and why? What pros/cons of each make the choice?

Daniel Newman (@danielnewmanuv) 's Twitter Profile Photo

I don’t think people fully appreciate the amount of demand that inference will create for AI infrastructure. Training trillion parameter models will pale in comparison to volumes of trillions of concurrent tokens performing 24/7/365 across every business, organization, and

Jen Gentleman 🌺 (@jenmsft) 's Twitter Profile Photo

Me: I'm gonna put this somewhere safe My brain: Got it. Somewhere safe. So safe. Safe even from you. I will strip the memory of the location from all caches. No one will find it Me: No wait

Andrew Jones (@hpcnotes) 's Twitter Profile Photo

Is teaching #programming in hand-coded #Fortran a bit like teaching assembly now? Should all programming teaching now be based on how to write software with intelligent use of #AI coding tools?

Daniel Newman (@danielnewmanuv) 's Twitter Profile Photo

When people are afraid of change and disruption, they dig into the past with all of their might while adamantly denying the future. There was an error where teachers wouldn’t let you use calculators in math because supposedly it was gonna slow down. Learning in math has evolved