Draft Correction: Post 29 - Serial Number One
Draft Correction: Post 29 - Serial Number One
Post: /_posts/2026-05-04-Serial-number-one.md
Verification date: 2026-05-17
Status: REVIEW COMPLETE
SERIOUS ERRORS (must be fixed before publication)
1. CDC-IBM antitrust settlement amount
Post says (paragraph “Coda”): “It survived a major IBM antitrust settlement in 1973 (CDC received six hundred million dollars)…”
Correct: The 1973 IBM-CDC antitrust settlement was approximately $80 million in cash and assets, including the transfer of IBM’s Service Bureau Corporation. Multiple sources (Wikipedia “Control Data Corporation”, TIME magazine archive of “Settlement for IBM”, and IBM antitrust documents) confirm the $80 million figure.
The post’s $600 million figure may be confusing this with the cumulative damages CDC sought, or with later IBM antitrust matters with the federal government, but the actual settlement received by CDC was about $80 million. This is a substantial error of magnitude (~7.5x).
Action: Change “six hundred million dollars” to “approximately eighty million dollars in cash and assets, including the transfer of IBM’s Service Bureau Corporation.”
2. CDC 1604 announcement venue: Western Joint Computer Conference, Los Angeles
Post says (“Building the 1604” section): “CDC publicly announced the machine on 16 October 1959 at the Western Joint Computer Conference in Los Angeles.”
Correct: The CDC 1604 release date of October 16, 1959 is correct. But the WJCC 1959 was in San Francisco on March 3-5, 1959 — not in October, not in Los Angeles. The Eastern Joint Computer Conference 1959 was in Boston, December 1-3, 1959.
No JCC was held in October 1959. The October 16 1959 announcement was most likely a corporate press release, not at a major industry conference. (Joint Computer Conference - Wikipedia.)
Action: Remove the WJCC / Los Angeles attribution. Replace with “publicly announced the machine on 16 October 1959” without the venue claim, or verify a specific announcement venue from CDC corporate sources.
3. Cray’s father was a civil engineer, not a city engineer
Post says (Cray bio paragraph): “He had been born on 28 September 1925 in Chippewa Falls, a small lumber town in northwestern Wisconsin where his father was the city engineer.”
Correct: Per the Wikipedia article on Seymour Cray, his father was a civil engineer (a profession), not a “city engineer” (a municipal job title). This is a factual misstatement. Some sources say his father was the Chippewa Falls city engineer, but Wikipedia explicitly distinguishes — Markoff’s NYT obit doesn’t use the “city engineer” framing either. The terms are not interchangeable.
Action: Change “city engineer” to “civil engineer” unless a primary source can be cited supporting the city engineer claim.
4. FNWF was renamed from NANWEP in 1961, not founded in 1959
Post says (paragraph 2): “Eight months later, in August 1960, the Naval Postgraduate School’s Fleet Numerical Weather Facility (FNWF), running on the same CDC 1604…”
More accurate: In August 1960, the unit was still called NANWEP (Navy Numerical Weather Problems group), which had moved to NPS Monterey in March 1959 under Captain Wolff. It was renamed Fleet Numerical Weather Facility (FNWF) only in 1961.
The post elsewhere (“Spanagel Hall, January 1960” paragraph) calls it “FNWF” throughout — this is anachronistic for the January-August 1960 period.
Action: Either (a) introduce the NANWEP-then-FNWF naming history when first mentioning the facility, or (b) refer to it as “Wolff’s group” or “the Navy numerical weather group” for the 1959-1960 period.
5. Cray Research startup capital: $250,000 vs $300,000
Post says (Coda): “He founded Cray Research, Inc. with three hundred thousand dollars of CDC’s seed money…”
Disputed: Wikipedia “Seymour Cray” gives the figure as $250,000 of CDC start-up money from Norris. Other secondary sources give $300,000. The post’s figure may be correct but is not the dominant figure in the most authoritative source (Wikipedia citing the Murray book).
Action: Either cite $250,000 from Murray/Wikipedia, or footnote the $300,000 with the specific source claim.
6. Cray-3 / Graywolf treatment
Post says (Coda): “Cray Research shipped… and the cancelled Cray-3 (Graywolf) in the early 1990s.”
Correction: The Cray-3 was NOT a Cray Research product — it was a Cray Computer Corporation product (CCC was Cray’s separate post-1989 company). Graywolf was the specific Cray-3 prototype delivered to NCAR on May 24, 1993 as a loaner/evaluation unit (one unit; never paid for; CCC went bankrupt in March 1995). The Cray-3 was not “cancelled” — one unit shipped. The Cray-4 was the cancelled follow-on design.
The post conflates the Cray-3 program with the Cray-4 cancellation, and incorrectly attributes the Cray-3 to Cray Research (it was CCC).
Action: Rewrite to: “Cray Research shipped the Cray-1 in 1976 and the Cray-2 in 1985. Cray himself left Cray Research in 1989 over a strategic disagreement on multi-processor design and founded a third company, Cray Computer Corporation, in Colorado Springs, which delivered a single Cray-3 prototype (named Graywolf) to NCAR in May 1993 and went bankrupt in 1995 before completing the Cray-4.”
SMALLER ERRORS / IMPRECISIONS
7. “Westinghouse-trained electrical engineer” for Norris
Post says (ERA: Out of CSAW section): “William Norris, a Westinghouse-trained electrical engineer from Inavale Nebraska who had also spent the war on cryptanalysis”
More accurate: Norris earned his electrical engineering degree at the University of Nebraska in 1932. He later worked for Westinghouse in Chicago selling X-Ray equipment (not as an engineer trainee in the standard sense). “Westinghouse-trained” overstates the Westinghouse role and undercuts the Nebraska training. Also, Norris was born near Red Cloud, Nebraska, and attended a tiny school in Inavale — describing him as “from Inavale” is technically fine, but combine this with “Westinghouse-trained” is misleading.
Action: Change to “William Norris, a Nebraska-born electrical engineer (B.S.E.E., University of Nebraska 1932) who had also spent the war on cryptanalysis” or similar.
8. “Cray was thirty-two when he joined CDC”
Post says (Cray and Little Character section): “Cray was thirty-two when he joined CDC.”
Check: Cray was born 28 September 1925. CDC was incorporated 8 July 1957 (when Cray was still 31), but Cray joined formally in late 1957 / early 1958 (post says he resigned from Sperry only after AN/USQ-17). If he joined CDC after September 28, 1957, he was indeed 32 (or 32 turning 32). This is plausible but worth re-checking against Murray.
Action: Probably fine; verify if any specific date can be cited.
9. Excerpt and intro “the first fully transistorised scientific computer”
Post says (excerpt): “It was the first fully transistorised scientific computer.”
More precise: Per Wikipedia, the CDC 1604 was “one of the first commercially successful transistorized computers” — not necessarily “the first” fully-transistorized. The IBM 7090 was also transistorized, announced earlier (November 1958), and other transistor scientific computers existed (e.g., Philco TRANSAC S-2000). The 1604 was probably the first commercial transistor scientific computer designed from the ground up by a startup, but the bold “first fully transistorised scientific computer” claim is overstated.
Action: Soften to “one of the first fully transistorised commercial scientific computers” or “the first transistor scientific computer designed by Control Data Corporation.”
10. “BESM-6… instruction set was modified during development to be partially compatible with Western (specifically CDC) software”
Post says (Soviet sale section): “the Soviet BESM-6 computer of 1968 was originally designed by Sergei Lebedev’s team in Moscow as an indigenous Soviet supercomputer, but its instruction set was modified during development to be partially compatible with Western (specifically CDC) software.”
Unverifiable: Wikipedia’s BESM-6 article contains NO claim of CDC/Western compatibility — only standard FORTRAN/ALGOL/Pascal compiler support. The “partial CDC compatibility” claim does not appear in primary public references; the post itself flags it as somewhat speculative (“partly speculative”). This needs strong citation or removal.
Action: Either cite Trogemann/Nitussov/Ernst (post does cite this as footnote 17), or soften further. Confirming the specific CDC compatibility claim from primary sources is required.
11. “WMIN… 250-watt radio operator”
Post says (the walkout section): Mullaney “had worked nights as a 250-watt radio operator at the Minneapolis station WMIN to pay for his tuition”
Status: Could not verify the WMIN call-sign claim or the 250-watt detail in the time available. This is a Mullaney CBI oral-history-level claim — likely true if from Mullaney himself, but could not be independently verified.
Action: Confirm against Mullaney CBI transcript. If true, fine; if not, soften.
12. National Medal of Technology year (1985 vs 1986)
The official year for Norris’s award is debated. National Medal of Technology winners for 1985 included some, and Norris is dated as 1986 by Reagan Library records. The post says “1986” which appears correct per Reagan Library.
Action: No change needed.
13. “Marathon Oil… Masquerade” text-mining application
Not verified. Could not find primary source for the “Marathon Oil Masquerade” 1604 text-mining application claim in the time available. Likely true (this is the kind of niche claim that would come from Murray’s The Supermen or CBI archive) but needs sourcing.
14. “Contemporary tube-era IBM 709” — comparison muddled
Post says (“Building the 1604” section): “The total power draw of the production 1604 was about 5.5 kilowatts, against roughly 50 to 60 kilowatts for the contemporary tube-era IBM 709.”
Imprecise: The IBM 709 was indeed tube-era (released August 1958). But the actual contemporary machine of the CDC 1604 was the transistorized IBM 7090 (announced 1958, first installation December 1959 — basically contemporaneous with the 1604). The 7090 had reduced power requirements (~70% reduction vs tube machines). Comparing 1604 power to the older 709 rather than the contemporary 7090 understates how competitive the IBM transistor machine was on power. The “contemporary tube-era IBM 709” phrase is internally contradictory — 709 was tube-era and not contemporary; 7090 was contemporary and not tube-era.
Action: Reframe the comparison: either (a) “against the tube-era IBM 709 (around 50-60 kW)” without “contemporary,” or (b) compare to the IBM 7090’s specs directly.
15. Date precision: Cray accident date
Post says (Coda): “On 22 or 23 September 1996 – sources differ by a day – Cray was driving his Jeep Cherokee on Interstate 25…”
More accurate: The accident occurred on Sunday, September 22, 1996, per UPI archives. The September 23 figure may come from sources that reported the next day. There’s no real “sources differ by a day” issue — the accident was Sunday Sept 22 and was reported by media on Monday Sept 23.
Action: Change to “On Sunday, 22 September 1996, Cray was driving his Jeep Cherokee…” Remove the “or 23” / “sources differ by a day” framing as misleading.
UNVERIFIABLE BUT NOT FLAGGED AS ERRORS
- The exact “two billion computations per analysis-and-forecast cycle” from the 1963 CDC sales brochure — could not verify the exact wording in the time available, but the brochure does exist (Weather By Computer, 1963).
- The “thirty-seven cents” germanium transistor anecdote — the post itself flags this as period anecdote not fully primary-source confirmed.
- The specific Soviet export-licence details (1604A to JINR vs CDC 6400 to Yerevan).
- The “five-year goal / one-year goal” Cray quote — the post says “well attested across multiple oral-history sources and Murray (1997).” This is likely true given Cray’s well-known reticence and the canon Murray drew on.
- “approximately seven hundred plug-in cordwood modules” in a 1604 — could not verify exact number.
- “1.2 million dollars committed, double the prospectus target” for the CDC IPO — could not verify against Murray in time available.
CONFIRMED-CORRECT SURPRISES
- The NCAR-1604 myth is correctly debunked in the post: NCAR’s first computer was the CDC 3600 in November 1963, NOT a 1604. This is good fact-checking.
- The “Cray was at Battle of the Bulge in Europe” correction (vs popular Pacific-only narrative) is correctly attributed to the Markoff NYT obituary, and matches the obituary text.
- CDC was incorporated 8 July 1957 — verified.
- CDC 1604 serial 001 delivered to NPS Monterey in Spanagel Hall Room 101A in January 1960, supervised by Cray — verified by Computer History Museum and NPS sources.
- Cray’s BSEE 1949 and MS 1951 at Minnesota (not Wisconsin as some sources claim) — verified.
- Engstrom (1902-1962) Yale PhD math 1929, CSAW WWII — verified.
- Norris died 21 August 2006 at age 95, Parkinson’s, Bloomington — verified.
MISSING FOOTNOTES / UNDER-CITED CLAIMS
- Footnote 12 acknowledges the thirty-seven-cents anecdote is hard to verify; good honest footnote.
- Footnote 17 cites Trogemann/Nitussov/Ernst for BESM-6, but the specific CDC-compatibility claim should be cited more precisely or softened.
- The CDC-IBM settlement amount in the Coda lacks a footnote and is incorrect (Error #1 above).
- The “501 Park Avenue + 1103 = 1604” naming origin — post notes the company’s own engineers said it was “probably apocryphal” and Wikipedia treats it as unverified legend. The post handles this reasonably.
- The Marathon Oil Masquerade claim lacks a footnote.