4 ms·

Can it run Doom?

retrac 5y ago
Well, Carmack took the question at least a little seriously. https://news.ycombinator.com/item?id=25494648 https://news.ycombinator.com/item?id=25494648 tl;dr: Quite probably. It's certainly fast enough, but it's an awkward beast without byte-addressed memory. (Also, there's the minor problem that there's no video hardware.)
- rbanffy 5y ago
  > Also, there's the minor problem that there's no video hardware. I believe there were a few frame buffers available. IIRC, the movie "The Last Starfighter" had their VFX rendered on an Y-MP and they would require some form of pre-visualization during the production process, but transfers from render to viz could be achieved with network or magnetic storage sneakernet.
  - ahefner 5y ago
    If there's anything halfway plausible that could've been connected directly to the Cray, this is a golden opportunity for someone to write the first demo on an exceptionally historic platform. Good luck ever seeing it running on real hardware, though.
    - watersb 5y ago
      In that era, rendering of the frame buffer was often performed by writing to a film slide recorder. That is, it streamed a bitmap to a printer that wrote to optical film stock. We had a slide recorder at one of my early IT jobs. IIRC, it was at output device for a Macintosh computer, which meant that it was technically my responsibility to keep running, but it was a special beast so I would basically check to see if it was plugged in correctly and not on fire; for anything else, I would call the vendor that leased it to us. I'm pretty sure that the film stock, once written, needed to be developed like any other photo film, but that was 30 years ago and I just can't recall. But sure, kid: write a demoscene for it. 0.0005 frames per second...
      - ahefner 5y ago
        I'm not sure that qualifies as halfway (or even remotely) plausible, but with sufficient pipelining (tens of thousands of optical printers with parallel IO hardware and an equal number of people developing slides and loading them up), miracles may be possible. Given resources on the scale of the Apollo program, you could make some super sick demos. Way cooler than collecting moon rocks. Joking aside, this essentially hardcopy-to-film is a thing that I never considered existing, yet sounds completely plausible. Even so, a color framebuffer with a CRT attached was certainly not unobtainium by the early to mid 80s.
        watersb 5y ago
        Absolutely, yes color frame buffers and CRT display was widely available. The Cromemco S-100 bus computers were used for television news weather visualization and for engineering applications. Hmm. I think perhaps the limit was on resolution and color depth. The movie "TRON" was done with slide recorders, and now I have another rabbit hole to dig out of! (The BYTE Magazine article that had me thinking of high end graphics at the time, and the use of film as output, is from November 1982. I was in 8th grade and read every page of these monstrous, 400-page times. Like the Sears Christmas catalog, every month.) https://archive.org/details/byte-magazine-1982-11/page/n49/mode/1up https://archive.org/details/byte-magazine-1982-11/page/n49/m... (I could just go ask the old Cray-1 that's a block up the street from here, but he is now a museum display and doesn't talk much.)
        rbanffy 5y ago
        > I could just go ask the old Cray-1 that's a block up the street from here, but he is now a museum display and doesn't talk much. Show off ;-) The first SGI IRIS shipped in 1983. You could probably stream a 320x200 8-bit bitmap to it via ethernet, but that'd probably be considered cheating because another computer holds the display. The grand dad of the X-MP, the CDC-6x00 had a couple peripheral processors that could push memory out (used in those awesome dual-round-screen terminals), and I remember seeing a terminal like those (with a single screen) running off a Cray-1 somewhere, so it could be possible with hardware actually used at the time.

Can it run Doom?

Well, Carmack took the question at least a little seriously. https://news.ycombinator.com/item?id=25494648 https://news.ycombinator.com/item?id=25494648 tl;dr: Quite probably. It's certainly fast enough, but it's an awkward beast without byte-addressed memory. (Also, there's the minor problem that there's no video hardware.)

rbanffy 5y ago
> Also, there's the minor problem that there's no video hardware. I believe there were a few frame buffers available. IIRC, the movie "The Last Starfighter" had their VFX rendered on an Y-MP and they would require some form of pre-visualization during the production process, but transfers from render to viz could be achieved with network or magnetic storage sneakernet.
- ahefner 5y ago
  If there's anything halfway plausible that could've been connected directly to the Cray, this is a golden opportunity for someone to write the first demo on an exceptionally historic platform. Good luck ever seeing it running on real hardware, though.
  - watersb 5y ago
    In that era, rendering of the frame buffer was often performed by writing to a film slide recorder. That is, it streamed a bitmap to a printer that wrote to optical film stock. We had a slide recorder at one of my early IT jobs. IIRC, it was at output device for a Macintosh computer, which meant that it was technically my responsibility to keep running, but it was a special beast so I would basically check to see if it was plugged in correctly and not on fire; for anything else, I would call the vendor that leased it to us. I'm pretty sure that the film stock, once written, needed to be developed like any other photo film, but that was 30 years ago and I just can't recall. But sure, kid: write a demoscene for it. 0.0005 frames per second...
    - ahefner 5y ago
      I'm not sure that qualifies as halfway (or even remotely) plausible, but with sufficient pipelining (tens of thousands of optical printers with parallel IO hardware and an equal number of people developing slides and loading them up), miracles may be possible. Given resources on the scale of the Apollo program, you could make some super sick demos. Way cooler than collecting moon rocks. Joking aside, this essentially hardcopy-to-film is a thing that I never considered existing, yet sounds completely plausible. Even so, a color framebuffer with a CRT attached was certainly not unobtainium by the early to mid 80s.
      - watersb 5y ago
        Absolutely, yes color frame buffers and CRT display was widely available. The Cromemco S-100 bus computers were used for television news weather visualization and for engineering applications. Hmm. I think perhaps the limit was on resolution and color depth. The movie "TRON" was done with slide recorders, and now I have another rabbit hole to dig out of! (The BYTE Magazine article that had me thinking of high end graphics at the time, and the use of film as output, is from November 1982. I was in 8th grade and read every page of these monstrous, 400-page times. Like the Sears Christmas catalog, every month.) https://archive.org/details/byte-magazine-1982-11/page/n49/mode/1up https://archive.org/details/byte-magazine-1982-11/page/n49/m... (I could just go ask the old Cray-1 that's a block up the street from here, but he is now a museum display and doesn't talk much.)
        rbanffy 5y ago
        > I could just go ask the old Cray-1 that's a block up the street from here, but he is now a museum display and doesn't talk much. Show off ;-) The first SGI IRIS shipped in 1983. You could probably stream a 320x200 8-bit bitmap to it via ethernet, but that'd probably be considered cheating because another computer holds the display. The grand dad of the X-MP, the CDC-6x00 had a couple peripheral processors that could push memory out (used in those awesome dual-round-screen terminals), and I remember seeing a terminal like those (with a single screen) running off a Cray-1 somewhere, so it could be possible with hardware actually used at the time.