Talk:Vacuum-tube computer

  This article is or was the subject of a Wiki Education Foundation-supported course assignment. Further details are available on the course page. Student editor(s): Ltorack, SeanLoreaux. Peer reviewers: Spike64113, Seth Listhartke, Rtthb.

Above undated message substituted from Template:Dashboard.wikiedu.org assignment by PrimeBOT (talk) 04:37, 18 January 2022 (UTC)Reply

My partner and I plan to add a new section to this article about the usage and contributions of vacuum based computing during the second world war. SeanLoreaux (talk) 16:14, 6 April 2018 (UTC)Reply

The article claims that "The ENIAC computer (1946) had over 17,000 valves and … consumed over 150 kilowatts of power". Although sourced, is the source correct? It works out at an average of nearly 9 watts per tube. That is the order of power of a medium sized power output tube. Assuming the majority are small signal tubes (for the logic) and are the typical 3 watts per tube (mostly the heater), that means that ENIAC would have to have had some very hefty tubes in it somewhere to run up that sort of power bill.

By comparison, Colossus had 2600 tubes but managed with a mere 8 kilowatts (or 3 watts a tube). DocFergus (talk) 12:47, 10 November 2018 (UTC)Reply

Pretty commonly cited number. A quick Google Books search on "Eniac Kilowatts" produces a passage in "Power-Efficient Computer Design" ^[1] which cites a press release saying that the ENIAC used 150 kw from a 240 V 3 phase supply, of which 80 kilowatts were used for heating tubes, 45 kilowatts for generating DC voltages, 20 kilowatts for ventilation blowers, and 5 kilowatts for punched-card machinery. Why do you assume they were small signal tubes? 80 kilowatts divided by 17,000 tubes gives more than 4 watts per tube for heaters, or about .8 amps at 6 volts - higher than the miniature receiving tubes I used to mess around with, but not unreasonably high. A discussion of the power consumption of the different vacuum tube computers would be fascinating, but over-specialized for this article. 5 kW of punch card machinery is also pretty impressive when I think of the Modle 029 I used in school which plugged into a wall socket and couldn't have used more than a couple hundred watts. --Wtshymanski (talk) 18:43, 10 November 2018 (UTC)Reply

@ArnoldReinhold, you have included several digital programmable calculators in the list which do not have stored program capabilities and do not fit the usually accepted definition of a computer. I will revert your addition of the 604 and CPC in the list for the same reason, but we can discuss this. Maybe we could add two sections, one for digital electronic calculators and the other for stored program computers, even though I think this article should concentrate on actual computers. However we shouldn't mix both of them in the same table as these are not the same kind of machines.

Also, the list of stored-program computer is still far from complete, many first generation computers were created in Europe in the early fifties and would benefit from being more known (Lyons Electronics and Eliott in the UK, SEA in France, Zuse in Germany...). Maybe the introduction to this section should reflect this.

Damien.b (talk) 18:55, 1 September 2023 (UTC)Reply

@Damien.b: This article is about vacuum-tube computers, not just stored program models. It states "The Atanasoff–Berry computer, a prototype of which was first demonstrated in 1939, is now credited as the first vacuum-tube computer." It goes on to discuss Colossus and ENIAC neither of which were stored program devices. The intro also states that vacuum tube computers were mostly one of a kind, which was true at first, but an entire generation of multi-copy commercial vacuum tube computers emerged in the 1950s.

Part of the problem is the over-broad term "first generation computer" which encompasses both the large numbers of one-off experimental designs and the first commercial machines, which created the computer industry we now know. We need overview content on the commercial vacuum tube generation, which has been ignored on Wikipedia, and this article seems the appropriate place.

The IBM 604 is as much of a computer as the Atanasoff–Berry and it sold in volume (5600 units). It is by far first large scale use of vacuum tubes for computation and belongs in this article. The IBM CPC combined the 604 with other devices, including a 10 word memory, to perform more advanced calculations, and hundreds were sold. Our article mentions the Bull Gamma 3, but that article says it was "Originally designed as an electronic accelerator for electromechanical tabulating machines, similar to the IBM 604, it was gradually enhanced with new features and evolved into a first-generation stored program computer (Gamma AET, 1955, then ET, 1957)." Of course we should distinguish between calculators like the 604 and stored program computers which my reverted edits attempted to do.

I have no problem adding other commercial vacuum-tube computers that have Wikipedia articles or coverage to the list. The key requirement should be commercial designs built and sold in some quantity, even a handful, and experimental one-offs. I think I have justified restore the edits I made and I will add the items you suggested.--agr (talk) agr (talk) 21:45, 5 September 2023 (UTC)Reply

I’d also call attention to this link [1] from our IBM 604 article which describes how the 604 at Columbia University was used to solve simultaneous linear equations, exactly the task the Atanasoff–Berry computer was designed for.—agr (talk) 03:41, 6 September 2023 (UTC)Reply

I understand your point, and agree with you that the word computer can have many definitions. However we already have an entire list of vacuum tube computers in its broader term: List of vacuum-tube computers

I guess we should then just redirect to this article if we're to include them all.

Damien.b (talk) 06:55, 6 September 2023 (UTC)Reply