Cray
Computer companyWikipedia
13
MENTIONS
8
EPISODES
8
PODCASTS
Search complete. 13 mentions across 8 episodes found for "Cray".
Sep 17, 2026
OMFG: Lending to the Other 100 Million Americans: QAV America #70
T
17:15Tony KynastonHOST
yeah, right.
C
17:17Cameron ReillyHOST
One of the founders of CDC was Seymour Cray, developed some of the fastest mainframes in the world.
C
17:23Cameron ReillyHOST
During the 1970s, he left and founded Cray Research and pretty much is the father of supercomputers.
C
17:32Cameron ReillyHOST
So, you know, pretty much all of the AI stuff that's going on today, you can trace back to Seymour Cray one way or another.
C
17:39Cameron ReillyHOST
I remember my...
T
17:40Tony KynastonHOST
Jensen Huang of his day.
C
18:43Cameron ReillyHOST
I'll come down and we'll build a tunnel under your new home when you move in.
C
18:48Cameron ReillyHOST
See if the elves come to us.
@HPCpodcast-110: John Gustafson on AI, HPC, Making Every Bit Count – In Depth
J
17:47John GustafsonGUEST
So one of the rules of the HPL benchmark based on LINPACK is that it has to use 64-bit representation of floating point.
J
17:55John GustafsonGUEST
It does not have to be a standard floating point because that benchmark has been around since the days of things like the Cray-1.
J
18:02John GustafsonGUEST
I've actually written a paper about this that's being refereed right now.
J
18:07John GustafsonGUEST
But the original benchmark was to solve a 100 by 100 set of equations with 64-bit floats and see how fast you could do that using just standard Gaussian elimination with pivoting.
J
22:31John GustafsonGUEST
And then the, because they really struggled to build a 48 times 48 integer multiplier for the significance.
J
22:38John GustafsonGUEST
And in fact, they did a sloppy job of it and left off the least significant bits.
J
22:42John GustafsonGUEST
As a result, on a Cray-1, you could take A times B and get a different answer from B times A.
J
22:47John GustafsonGUEST
Interesting.
Head-ON With Roxanne Kincaid, 14 September 2026, Moran Monday
T
91:59ToddAUDIENCE
Maybe the servers, because as far as I understood it, there were some special chips that had to be really cool.
T
92:06ToddAUDIENCE
Like, you know, I remember the Cray supercomputer and all that.
T
92:09ToddAUDIENCE
Apparently, these quantum computers need to be cool, but things keep getting miniaturized.
T
92:18ToddAUDIENCE
And she isn't supposed to be making me think about it.
The Silicon Gold Rush: How AI is Driving the Development of New Chips
B
9:07Bill DallyGUEST
... that they have is they have, they have a, they have a pretty neat interconnection network with the, uh, with the training TPUs of the 3D torus.
B
9:13Bill DallyGUEST
I built a lot of supercomputers, you know, both at MIT and with Cray in, in the '90s, with 3D torus networks-
D
9:19Dave PattersonMODERATOR
Yeah
B
9:19Bill DallyGUEST
... and so I'm, I'm very fond of that, and I can see the hand of the co-author of my book, Brian James-
The Hard Lessons Behind Computing’s Biggest Bets
G
16:32Glenn LockwoodGUEST
Cori, like Gordon, was a super cool, very innovative system.
G
16:38Glenn LockwoodGUEST
Cori was a Cray XC30, I want to say.
G
16:42Glenn LockwoodGUEST
But it was the first large-scale deployment of Intel's Knights Landing processor.
G
16:47Glenn LockwoodGUEST
I'm sure you remember that and how exotic it was.
N
19:00NicoleHOST
blast from the past there.
G
19:03Glenn LockwoodGUEST
Yeah, we've talked about this plenty in its heyday in the mid-2010s.
G
19:11Glenn LockwoodGUEST
But Cori was, it shipped with this burst buffer technology called DataWarp that was co-developed between NERSC, Los Alamos National Lab, and Cray.
G
19:22Glenn LockwoodGUEST
And DataWarp was designed to solve this problem that Flash was too expensive to put everywhere.
#560: Building a Research OS: From Django to 30,000 Samples
M
30:44Michael KennedyHOST
What a different time it is for universities, and I think w- they're going through a, a similar upheaval, I guess, is the word.
M
30:51Michael KennedyHOST
When I was in university, and we're talking, like, '90s, there was a giant Cray supercomputer or maybe a, a, a couple of supercomputers at the heart of the university in some basement, and you could get access to that if you needed mega computing resources.
M
31:06Michael KennedyHOST
And now it's just, you know, scaled into some ginormous cloud infrastructure, and the limit is really just how much are they gonna give you, not what, what does the university have in its basement or something like that, which is really interesting.
S
31:19Shaun ChuahGUEST
Well, uh, I mean, there are still some universities that are trying to build their own clusters.
Bite: The Machine that knows Everything [April 1985]
A
0:40Artificial IntelligenceNARRATOR
The machine that knows everything.
A
0:42Artificial IntelligenceNARRATOR
T.T.M.O. unveils the Crayon Telecluster.
A
0:45Artificial IntelligenceNARRATOR
A room-sized oracle that files away every encyclopedia ever printed and answers back in plain English.
A
0:51Artificial IntelligenceNARRATOR
The hardware, TT, M, O's Cray Intelli cluster is a complete state-of-the-art solution that allows you to discover everything known to man.
El Niño is coming
D
11:04David LeitchHOST
And when I was reading about the weather forecasting system, it seems that it's making quite a difference in weather forecasting.
D
11:11David LeitchHOST
The-- I believe the Bureau used to run its system, or used to run s- has a supercomputer, a Cray supercomputer, which as it turns out is quite energy intensive [chuckles].
D
11:22David LeitchHOST
And that, uh, the Europeans and Google and everyone have, have tried to apply AI machine learning to weather forecasting.
D
11:31David LeitchHOST
And since May, uh, our weather forecasts also make use of such a system.