I am using SGX530 in TI DM8168. In my application, the build in function “smoothstep” would be called repeatedly. I’d like to know how many clocks that SGX530 needs for this function? suppose this smoothstep would be called 1000 time, taking the instruction loading into account, how many clocks would be needed to finish it?
Our offline GLSL profiling compilers give per-line cycle count estimates and a simulated total cycle count that goes through the same optimization passes as the online compiler (so should be almost identical to the actual on device cost). You can download the compilers from here.
Viewing 2 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic.