allocating shared memory...

Relevance of shared memory bank conflicts in Fermi and higher...

Effective Bandwidth in Cuda...

Strategy for minimizing bank conflicts for 64-bit thread-separate shared memory...

configure local (shared) memory for OpenCL using Nvidia platforms...

CUDA efficient usage of shared/local memory?...

Local, global, constant & shared memory...

Is it worthwhile to pass kernel parameters via shared memory?...

Kernel launch failure if the amount of shared memory allocated for the whole grid exceeds 48kB...

GPU shared memory practical example...

How to define a CUDA shared memory with a size known at run time?...

Does reading an int array from shared memory preclude bank conflicts?...

Will the same thread accessing the same memory bank twice cause conflicts?...

Is there a limit to OpenCL local memory?...

Can my kernel code tell how much shared memory it has available?...

Using both dynamically-allocated and statically-allocated shared memory...

Maximum (shared memory per block) / (threads per block) in CUDA with 100% MP load...

Kernel lauch specifying the stream, but with default shared memory size...

cudaFuncSetCacheConfig - applies to all devices or current device?...

How can I launch a kernel with "as much dynamic shared mem as is possible"?...

How come my kernel doesn't exceed the shared memory limit?...