occupancy means in GPU programming

CPU cache

L1/L2 cache hierarchy reduces global memory latency

warp divergence kills performance

Warp divergence causes threads to execute non-uniformly, leading to idle cycles and reduced throughput

tensor cores are

Why can computers crunch numbers faster than humans?

CUDA

CUDA enables parallel computation on GPUs

Dynamic random-access memory

DRAM requires periodic refreshing to maintain data integrity

register pressure means: too many variables per thread reduces occupancy

Too many variables per thread reduces occupancy