For quotes by others regarding Seymour and his career click here

Verified Seymour Quotes

There is little proof that Seymour *actually* said most of the things that have been attributed to him. There are *lots* of unverified quotes out there on the internet, these are the ones that seem the most like something that Seymour would have actually said.

The following quotes have been verified, i.e., there are actual video or audio tapes that provide a record of what Seymour said.

"People buy these big computers for emotional reasons, it's not that they really need them at all."

- 1985, videotaped interview in Chippewa Falls, WI.

Unverified Seymour Quotes

Where we know of a reference, we list it. The following UNVERIFIED quotes have been widely attributed to Seymour R. Cray. There are many different versions of these quotes on the internet, along with many others that just don't seem plausible.

We'd like to hear from anyone who has additional information on the authenticity of any of these quotes to help set the record straight.

"Parity is for farmers."

- This quote is usually seen as Seymour referring to the original Cray-1 serial 1, which was delivered to Los Alamos Scientific Laboratory on a six month trial. However, in his excellent book A Few Good Men From Univac, author David E. Lundstrom recounts the story as Seymour speaking to a visitor to the Control Data Arden Hills lab who asked why Seymour didn't design memory parity into the original CDC 6600. A sometimes seen variant of this quote is "Parity is for dirt farmers".

"Farmers buy a lot of computers."

- This quote is attributed to Seymour when he was asked why he subsequently agreed to put error-correction into the Cray-1, beginning with SN3 which went to NCAR. Starting with SN3, the Cray-1 frame was approximately 8 inches taller, to accommodate the extra modules and circuitry required to add Single-bit Error Correction/Double-bit Error Detection (SECDED) to the Cray-1 memory architecture. Cray-1 SN2 was scrapped out.

It was Les Davis who led the team that implemented SECDED in SN3 that went to NCAR, so Seymour could continue on with his work on the Cray-2.

"If you were plowing a field, which would you rather use? Two strong oxen or 1024 chickens?"

- Comparing fast single processors to massively parallel systems. Seymour later settled on a design incorporating Intel microprocessors into his first system at SRC Computers, but he died shortly thereafter before his design could be realized.

"Anyone can build a fast CPU. The trick is to build a fast system."

- Speaking on the importance of balancing memory, bandwidth & throughput.

"Virtual memory leads to virtual performance."

- On why Seymour refused to offer virtual memory in his designs. This design limitation forced programmers to think about ways to optimize the performance of their software, which in turn optimized codes and reduced I/O.

"Memory is like sex, it's better when it's real."

- this one seems out of character for Seymour, the above quote seems more likely.