Cray1 Anniversary Event September 21st 4549

I was broached with the question in May. The answer came back, oh it was delivered in March. Too late. Yes, it is ironic. And I am not clear if the staff knows (well Al does and Lee does NOW, because they are reading). It's happening now because this is the available time. I've run into a bunch of friends of the speaker who can't attend and I am sneakernetting messages and cards, etc.

Why: it's even arguable that I visited Convex in Plano because of Cray. Several times.

Well, he was driven. I do know that when he was end, so believe American supercomputing if not the world also died (and all else are less thans). I also do know a few wanted violence against the driver who did it. I had a bright friend end in a similar way. At least Cray didn't go in pain. I was asked to act as intermediary to bring either Seymour or Chen to Hackers, and they seriously considered it. I can't post details. But like all hardware engineers, he had his set of faiures.

Last Friday I spent the last 4 hours interviewing my old office mate who knew him and introduced me to him. And I was asked to rush the interview for this meeting. In some ways Cray did somewhat sloppy architectures to get speed (you might be aware of why (a quote) he did 1-s complement for so long). He never adopted caches (I could have helped there). I did pump out of my officemate unconfidential details not only of the Crays 3 and 4, but also the Cray 5 and Cray 6 (how close the SRC 6 is to the Cray 6 I am not certain). He had a closer working relationship with Intel than I think most people knew.

The last time I looked at this it had a lot of typos. But then so do many of my own posts. And I think there may have been 1-2 technical misunderstandings.

--