Seymour Cray -- Father of the Supercomputing Industry
Вставка
- Опубліковано 5 вер 2024
- A Cray Research-produced documentary on Seymour Cray, the "Father of the Supercomputing Industry." Seymour Cray was the founder of supercomputer manufacturer Cray Research. Cray has been credited with creating the Supercomputer Industry.
"Be comfortable with yourself." Huge key when doing quiet creative work. That's something so important to work on. I keep chipping away at that and over the years have seen big results.
He seemed to change his mind a lot. Opposed to parallel computing (multiple instruction streams, not just SIMD) in favor of a fast single core, he later embraced parallel computing. Also ridiculed virtual memory as only for farmers, then doing a reverse, saying farmers buy a lot of computers.
I like that. It seemed he strived for simplicity & elegance, until technological reality made an approach no longer feasible.
It wasn't so much that Seymour Cray changed his mind as the technology changed during his long career. When he made the Cray I, other companies were making highly parallel computers for advanced scientific computations as well. But he made sure that his design would also be fast for problems that couldn't be divided into parts that could be done in parallel, which is why it was much more effective than the competition.
@dippywatcher I don't know that a 7600 ever ran NOS - only Scope 2. The 6000, Cyber 70 and Cyber 170's all did run NOS (Kronos) and NOS/BE (Scope 3). I am even trying to remember if the Cyber 176 ever ran other than Scope 2. I did work on them all under pretty much all operating systems both for CDC and clients.
They were probably the most fun systems I have worked on in a 40+ year career.
To answer a question, the Cyber 176 did have a NOS version, but NOS never supported the 7000 FLPPs, only the slower 6000 PPs. I was part of the grander porting program in 1979 while working at CDC.
Thank you, so much.
I love the tangible feeling of power you get when you look at a Cray machine.
I think an engineer like Kelly Johnson is also "up there". Made many contributions, but his SR71 was so mindblowingly ahead of it's time when introduced way back in the 60's, and it's crazy that even today it hasn't been beat in terms of perfomance (and cool looks).
Cray, Gordon Moore, Turing, Shockley: they created our present.
Shockley? No. Bardeen and Brattain. That well-known photo of the three together shows Shockley sitting at Brattain’s lab bench. Shockley sat there because he was the boss. It was self-promotion.
..."That's enough." Seymour Cray, what a fantastic mind.
Seymour Cray was and is the embodiment of American ingenuity.
How is it possible to "dislike" this?
amen zwa that shit cray
@Raymond Davies Ah, that all make sense, now. I gotta get out of the computer room and go fishin', every now and again. :)
No they are all as important as each other. They all need each other. We should be thankfull for all who participated in making computers faster BUT ALSO more affordable,accessible and useable by the ordinary guy.
love the video really good
6:11 mins of inspiration right there.
Thanks for sharing, what a remarkable man!
@nightowl8936 I'm pretty sure that the world(the people who know their stuff, anyway) knows of Cray's great contributions to computing. Of course, comparing a modern-day GPU to a Cray-3 or 8600, or whatever is silly. Regardless, the truth is that a HD 5870 is a great deal more powerful than a Cray-3/4, but that's to be expected with how fast technology moves forward. I'm not sure what's the big deal here.
Seymour cray is a scientist or engineer?
Give that man a medal
@nightowl8936 Calling people fools and idiots is, of course, the best way to prove you're right. The end of the Cold War didn't end the need for supercomputers, it merely changed the applications of supercomputers.
Though I'm not really sure what are we even arguing about?
awwwww dats soooo touching
@nightowl8936 Yes, I know that Cray was a great and talented man, I never said he wasn't. His contributions to computing were massive, but his ideas were becoming increasingly outdated by the 90s and his supercomputers did not see much success at that point.
Beyond my thought? Maybe. After all, Cray did claim that he spoke to elves in a tunnel that he dug beneath his house.
@nightowl8936 He's right though.
@gnossticc yeah.. I kinda realized that . and in 9 months I kinda got respect for apple.. so.. I think I'll just delete my stupid ass comment
Seymour wasn't perfect. His 160 computer had 64 instructions, but no subroutine call instruction! A BIG oversight. And I/O instructions all blocked. And there was no memory extension possible. But they were patched up as the 160A. The 6600 also had some troubling shortcomings-- If a peripheral processor hung, there was no way short of a complete computer deadstart to get it going again. I/O instructions still blocked. There was no good and simple way for processes to share any common code libraries.
On the other hand he did some genius things. Allegedly he wrote all of the Chippewa Operating system by hand, in octal, even the FORTRAN compiler. CDC eventually junked their own operating system, worked on by around 100 programmers in Sunnyvale, and just renamed COS as SCOPE. So one guy did what 100 couldn't.