AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Importance of Fusion in Parallel Execution
It's just not a good idea to split things into single operations. So the array languages that are doing this sort of fusion are are doing other stuff as well to try and keep some of the array power. I mean so if tooth arcs working on the GPU which is a very different model. And there you know splitting splitting a string operation on theGPU into sequential processing means, you know, I'll just ignore the fact that I'm running on a GPUs at all. And pretend it's a CPU and then get worse performance.