Brilliant move by Microsoft. You want to be the one selling shovels during a gold rush, as they say?
back
1 comments
nvidia is the shovel seller
quibble, but nvidia is the shovel maker. LLMs require a lot of vram but the underlying hardware requirements are actually fairly simple. There are already efforts in place to create inference specific ASICs.
That'll absolutely eat into Nvidia's profit margins.
Yeah computation has more layers and sellers, it's not a perfect analogy but good enough I feel