I wonder what "compute" would mean if CPUs were more efficient at matrix multiplication and vendors had the balls to pair each core to its own dedicated DDR and a star interconnect between.
It could be like the connection machine or something like that.
There is plenty of matrix multiplication in SIMD, but it isn't widely explored.
It could be like the connection machine or something like that.
There is plenty of matrix multiplication in SIMD, but it isn't widely explored.