Robert Crovella questions

Robert Crovella answers

• Share large constant data among cuda threads
• Independent Thread Scheduling since Volta
• Problem with determinism : set a buffer size in the CUBLAS_WORKSPACE_CONFIG environmental variable
• CUDA illegal memory access was encountered
• cuda SIMD instruction for per-byte multiplication with unsigned saturation
• CUDA placement new and virtual functions
• Thrust CUDA find maximum per each group(segment)
• Should I check the number of threads in kernel code?
• Does 'code=sm_X' embed only binary (cubin) code, or also PTX code, or both?
• Is it possible to execute multiple instances of a CUDA program on a multi-GPU machine?