3 ms·
i have a similar apocryphal story from the dod. there was an image processing element for a missile targeting system that was a 3d stack of silicon. it ran at a
by convolvatron 10y ago
i have a similar apocryphal story from the dod. there was an image processing element for a missile targeting system that was a 3d stack of silicon. it ran at an amazingly large number of ops per second, but only for a few seconds, because there was no way to get the heat out of the stack.
which was (apparently) fine, but it made testing very expensive
- i336_ 10y agoWhat year (or decade) was this, and what ballpark of OPS did it run at? That's amazing. Was the device able to be downclocked for certain tests?
- convolvatron 10y agothis was early 90s, i don't remember the ops, but i remember it being comparable in magnitude to the CM-2, which was where we were running some kind of simulation (in a skif, so everything I heard was very second hand)