Nvidia reportedly no longer supplying VRAM to its GPU board partners in response to memory crunch — rumor claims vendors will only get the die, forced to source memory on their own
https://www.tomshardware.com/pc-components/gpus/nvidia-reportedly-no-longer-supplying-vram-to-its-gpu-board-partners-in-response-to-memory-crunch-rumor-claims-vendors-will-only-get-the-die-forced-to-source-memory-on-their-ownOpen linkView original on lemmy.world
https://lemmy.today/post/42574307
If said rumor is true, so much for GPU prices falling.
EDIT: Well, I guess more properly, for video card prices falling; in this context, distinguishing between the GPU chip and the card it lives on does actually matter.
If board partners now provide their own VRAM, maybe they will provide more memory configurations for GPUs instead of the ones NVIDIA forces you to have.
In China street markets you can get a 48gb 4090.
I'm not sure if equipment to make older ram types still exist, or if it's possible to use ddr3 chips on ddr5 interface/voltage, or gddr5, but this is getting too stupid.
HBM prices are also rising fast, and next gen datacenter AI cards are likely to be worse value (flops/bandwidht/ram) than older ones. But still ddr5 would be more profitable to make than HBM at half the current price. So everything is stupid extortion, and should come back to some parity between the 2.
The entire pc market is about to rocket off once again... And just at the moment Windows has been a bitch about Windows 10 support. Coincidence?
If you think a secret cabal has been planning for the AI launch over ten years in advance (when the end date of Windows 10 was set), I have some other things to sell you.
We need to get off our nvidia dependency