Scratch Pad memory vs shared memory

Dear all:

What is the difference between a scratch pad memory used by an embedded processor and shared memory used by a GPU(NVIDIA)?


Not much, physically, they’re both small chunks of on-chip SRAM and can be used as user-managed cache. Scratch memory on scalar processors is typically only accessed by a single thread, whereas GPU shared memory is accessible by all threads in a given thread block and can be used for communication across threads.