News "Aurora’s Troubles Move Frontier into Pole Exascale Position" - HPCwire

[moinmoin]

With Sapphire Rapids delayed Intel's Aurora exascale supercomputer misses yet another date. This means Frontier is now on route of becoming the first exascale supercomputer.

Aurora's Troubles Move Frontier into Pole Exascale Position

Intel’s 7nm node delay has raised questions about the status of the Aurora supercomputer that was scheduled to be stood up at Argonne National Laboratory next year. Aurora was in the running to be the United States’ first exascale supercomputer although it was on a contemporaneous timeline with...

www.hpcwire.com

 

[DrMrLordX]

Intel is apparently using TSMC N7 and N5.

Minor quibble, but I think it's N6 rather than N7. Intel isn't getting any N7 wafers.

 

[Joe NYC]

Minor quibble, but I think it's N6 rather than N7. Intel isn't getting any N7 wafers.

It could very well be N6. I was just reading one of the recent articles that mentioned that the the tiles of Ponte Vecchio will contain N7 and N5, but it could very well be N6 instead of N7.

 

[moinmoin]

Inside view of the Aurora troubles, long twitter thread. (Mostly complaining about Intel failing scientific efforts by the repeated delays.)

https://twitter.com/x/status/1575233618170982400

 

[Tech Junky]

wow, almost a year since last post!

 

[Kocicak]

With such a delay, is the institution buying this computer even oblidged to still buy it? Cannot they cancel the deal and get their money back, if they already payed something?

 

[jpiniero]

With such a delay, is the institution buying this computer even oblidged to still buy it? Cannot they cancel the deal and get their money back, if they already payed something?

They basically got it for free. Intel had to pay a rather substantial penalty.

 

[DrMrLordX]

Inside view of the Aurora troubles, long twitter thread. (Mostly complaining about Intel failing scientific efforts by the repeated delays.)

Ouch.

It puts some things into perspective. Yes the Sapphire Rapids delays are as bad as any of us critics have said that they are. But, most of us in the peanut gallery will never be personally affected by said delays as much as are she and her colleagues in the hard sciences.

 

[moinmoin]

With El Capitan en route to be installed this year it seems it's time for a change of heads.

Alan Lee, at AMD for 15 years, founder of AMD Research and head of that as well as Extreme-Scale Computing Technology (so responsible for Frontier and El Capitan), is leaving:

Alan Lee on LinkedIn: AMD Research | 93 comments

After 15 years as the founder and head of AMD Research, it is time for me to move on. Over the years, AMD Research has created a large number of new… | 93 comments on LinkedIn

www.linkedin.com

AMD Research will be headed by Ivo Bolsens who joined AMD with the merger of Xilinx where he was SVP and CTO:

Head Of AMD’s Influential Research Group Leaves Amid Other Exec Changes | CRN

Alan Lee, the head of AMD Research, is leaving the chip designer amid other recent executive changes.

www.crn.com

He was CTO at Xilinx since 2001:

ivobolsens - Bio of Ivo Bolsens

5914 Porto Alegre Drive San Jose 95120 California

sites.google.com

Recent interview:

Looks like in a way Xilinx people are taking over.

 

[DrMrLordX]

Looks like in a way Xilinx people are taking over.

Well let's face it, if you've had a wild ride like being at AMD for the last 15 years, now might be a great time to retire. Take your fat paycheck/bonus that you've gotten for finally leading AMD out of the muck and relax.

It's sad to see someone who is at least indirectly responsible for their miraculous rise from near-bankrupcy fly the coop, but at the same time he was also there through some pretty dark times and at least indirectly responsible for some real goose eggs like Bulldozer.

Let's hope Bolsens knows what he's doing.

 

[A///]

It's sad to see someone who is at least indirectly responsible for their miraculous rise from near-bankrupcy fly the coop, but at the same time he was also there through some pretty dark times and at least indirectly responsible for some real goose eggs like Bulldozer.

Most people in that position with that power are very humble people. This industry, especially at the x86 level where everything is a measuring contest for those outside of the bubble looking in egos have no place in the workplace. It's all team effort mate. Without cohesion everything falls apart. All the teams under every division and group work together to bring advancements. Most people outside of the field and the adjoining hobby will have never heard of who people like Ivo or Jim are. They might know who Moore is, and only because the press has used his name for decades to explain his theory.