The memory bandwidth limitation is baked into the GB10, and every vendor is going to be very similar there.
I'm really curious to see how things shift when the M5 Ultra with "tensor" matmul functionality in the GPU cores rolls out. This should be a multiples speed up of that platform.
I'm really curious to see how things shift when the M5 Ultra with "tensor" matmul functionality in the GPU cores rolls out. This should be a multiples speed up of that platform.