map a PTX code block to GPU?

Accelerated Computing CUDA CUDA Programming and Performance
1

We are designing a compiler whose output is PTX assembly language which is generated from ANSI C ,my question is that how can this output be executed by GPU? In another word, how can we transmit this PTX code block to hardware?

2

Check out page 48 of the 2.2 Programming Guide. ;)

You’re interested in the cuModuleLoadDataEx() function in the driver API.