Talk:Seymour Cray

Archives (Index)

2014 2015 2016 2017

This page is archived by ClueBot III.

Weddings and children

The article currently says Seymour had one son and two daughters with Geri Harrand and cites https://www.independent.co.uk/incoming/seymour-cray-obituary-5595675.html which does say that and is a reliable source But I think it likely that the Independent article got it wrong, since the book The Supermen: The Story of Seymour Cray and the Technical Wizards behind the Supercomputer clearly says that he was previously married for almost 30 years to a Verene Voll, so it is more likely that he had the children with her instead. Unfortunately that book says little about his wedding, so it is hard to confirm from there. https://www.encyclopedia.com/humanities/encyclopedias-almanacs-transcripts-and-maps/cray-seymour-roger says that they divorced in 1975, and then he remarried Geri in 1980, who already had 3 children of their own. Unfortunately that article does not cite sources, but I think that article is likely correct. I'm sure that digging into other books would clarify this. Cirosantilli2 (talk) 16:08, 1 January 2022 (UTC)

Citation 38 page number is incorrect

Read the copy of where the link goes and page 205 made no reference of wedding only of the gallium circuits and silently firing a Vice President of engineering and purchasing production lines with stocks 2001:569:F99F:5700:CDE5:AD7D:BC72:8B6D (talk) 08:50, 19 January 2022 (UTC)

Suggested Improvements and Trivia

CDC's first computer was the CDC 160, followed by the 160A. These small machines sold commercially, and were used by CDC to help design the CDC 1604, and were used as I/O processors. I know of at least CDC 160 user that wrote a FORTRAN compiler for it. After initial deployment and development, the CDC 1604 used the then-common large computer overall architecture for batch processing, seven tape drives. Card decks from individual users were consolidated and processed as data by a CDC 160A, which produced a user FORTRAN input tape and a user program input data tape. The FORTRAN tape was processed by the CDC 1604 FORTRAN compiler to produce object code, which was written onto another tape. The object code was then loaded and executed. User data tape was on Tape 5. User program output was on Tape 6, a numbering that still survives as Fortran I/O default unit designations for input and output. Tape 6 was processed by another CDC 160A to produce output to line printers.

The CDC 6600 used a similar overall architecture, including use of CDC 160As to process input and output tapes for the CDC 6600.

The CDC 3000 series were self-contained. The computer read in user card decks, compiled and executed them, and produced line printer output as it compiled and executed each user job.

The CDC 6600 computers did use state-of-the-art hardware. A relative worked at Texas Instruments at the time and sold CDC the diodes used in the discrete logic; CDC did not use ICs for logic for the 3000 and 6000 series computers because chips were too slow then. My relative said that the speeds required made diodes difficult to produce, and to watch out for reliability problems with the CDC 6600. A year or two later, I used one of the early CDC 6600s and at the facility that I used, it was the most reliable computer that they had yet used throughout its lifetime of several years.

The CDC 1604 had the name CADET appended to it. This seems to have been an acronym, "Can't Add, Doesn't Even Try" because there was no arithmetic hardware and arithmetic was done using lookup tables.

The 1604, 3000 series and 6000 series all use six-bit characters (no lower case letters in the base character set). The 300 series had 24-bit words, and 48-bit extended words. Floating point hardware for 24-bit floating point was included. The CDC FORTRAN compiler included software double precision. I personally corrected that package, improving its speed and accuracy (the lower 16 bits were noise in the original package), improved the library sine/cosine packages by removing blunder and its patch in the CDC library code, and produced an all-new triple precision package for the CDC 3200. -- motorfingers : Talk 19:13, 28 April 2022 (UTC)

Regarding CADET: I think you are thinking of the IBM 1620. I've seen this moniker applied to the IBM 1620 in various places, including http://www.columbia.edu/cu/computinghistory/1620.html. Riordanmr (talk) 17:46, 18 February 2024 (UTC)

A Commons file used on this page or its Wikidata item has been nominated for deletion

The following Wikimedia Commons file used on this page or its Wikidata item has been nominated for deletion:

• Seymour Cray-crop.jpg

Participate in the deletion discussion at the nomination page. —Community Tech bot (talk) 21:22, 28 November 2022 (UTC)

Separate page for Cray Computer Corporation

I think Cray Computer Corporation requires its own page dedicated to company's history. Comrade-yutyo (talk) 18:10, 9 December 2022 (UTC)