4 ms·The closest I can think of is shared memory in CUDA / local memory in OpenCL.by twtw 8y agoThe closest I can think of is shared memory in CUDA / local memory in OpenCL.