cvi

I have found a way to increase speed in this situation. I wrote a multi-threaded version of memcpy, splitting the area to be copied between threads. Here are some performance scaling numbers for a set block size, using the same timing code as found above. I had no idea that the performance, especially for this … Read more