Talk:Seymour Cray/Archives/2015
![]() | This is an archive of past discussions. Do not edit the contents of this page. If you wish to start a new discussion or revive an old one, please do so on the current talk page. |
Chickens and Oxen
With reference to the section:
- Cray had always resisted the massively parallel solution to high-speed computing, offering a variety of reasons that it would never work as well as one very fast processor. He famously quipped "If you were plowing a field, which would you rather use: Two strong oxen or 1024 chickens?" By the mid-1990s this argument was becoming increasingly difficult to justify, and modern compiler technology made developing programs on such machines not much more difficult than their simpler counterparts.[citation needed]
Back in the real world, automatic parallelization in modern compilers has, for all intents and purposes, been a complete failure. The statement doesn't need citation, it needs deletion. — Preceding unsigned comment added by 171.64.172.206 (talk) 23:45, 19 January 2012 (UTC)
- Automatic parallelization may not have come to much, but it is true that practically all current supercomputer architectures are of the "1024 chickens" variety, rather than the "two strong oxen". Regards, Letdorf (talk) 22:56, 20 January 2012 (UTC).
Sigh. The MP-64 was going in that direction. When I left Silicon Valley, I posted a ( then new ) internal picture of an Intel 386 die on my office door at Cray Research and said " This will be our competition ". Work on a massively parallel follow-up to the multi-processor Y-MP was begun. We hired and I worked directly with Marv Bausman ( he holds several patents from the fiber optic clock we made. Note that my co-workers Ed Priest and Doug Paffel are names on that patent. ) and had most of the bleeding edge technologies solved when JR started cutting our funding and Seymours funding too.
I'm hardware. Chris Hsung was software and hardware architecture. Skip was " are we able to make it? "
Resonant box
The article says that the Cray 2 runs at 250MHz, or 4ns cycle time. A story I remember from a Cray seminar in the 1980's is that the Cray 2 was supposed to run at 4ns in a resonant box, but that it actually ran at 4.2ns, and the box had to be redesigned, as a 4ns box isn't at all resonant at 4.2ns. As far as I know, the 250MHz is wrong. Gah4 (talk) 20:11, 11 September 2015 (UTC)