@Eldeberen@social.middleearth.fr
2025-12-18 11:56 UTC
@ScriptFanix@maly.io @mike805@noc.social @MostlyHarmless@thecanadian.social You may be able to make waffer-scale chips, ignoring the defective areas/chips on it. That's almost what Nvidia do for its GPUs : high-end have a lot of working cores, low-end are the same but with more defective cores.
*But* for RAM this make so sens has it must be put along some computing chips to be useful, and I don't think the wafers bought have been design in that way…
And with waffer-scale chips you need to deal with a lot more issues such as heat management, dedicated encloruses (won't fit in a standard server) and so on
Replies (1)
-
@mike805@noc.social 2025-12-18 16:46
@Eldeberen@social.middleearth.fr @ScriptFanix@maly.io @MostlyHarmless@thecanadian.social Maybe they have some sneaky way to pair a memory wafer with a GPU wafer or portions thereof. For example, slice them both into quarters and put two memory quarters and two GPU quarters side by side on a heat absorbing substrate, and bond wire them together. Or they have something that can be put between two wafers to interconnect them in a stack. That would be even denser. OAI is going to pull some sort of hardware innovation out of a hat. They need one