GPU Computing for AI

Shared Memory

Fast, low-latency memory space shared among threads within the same block on a GPU, enabling collaboration and reducing accesses to the much slower global memory.

← Terug