gmemon questions

• SLI for multiple GPUs
• Synchronizing on an object and changing the reference
• CUDA-capable device
• L1 cache persistance across CUDA kernels
• Are indexes good or bad for a large database?
• Assigning random value to a parameter in a python program
• Root access required for CUDA?
• Multiple host threads launching individual CUDA kernels
• thread.start_new_thread vs threading.Thread.start
• Stop twisted reactor on a condition

gmemon answers