What steps will reproduce the problem?
1. Compile MD
2. Run MD on a CUDA witch 1.3 arch support
What is the expected output? What do you see instead?
Instead of the correct program output, we see the following message:
{{{
Cuda error: memcpy: too many resources requested for launch.
}}}
Possible solution:
Calculate the occupancy when parsing the file (like CUDA Occupancy
Calculator spreadsheet)
Original issue reported on code.google.com by namiurtf on 22 Apr 2010 at 12:38
Original issue reported on code.google.com by
namiurtf
on 22 Apr 2010 at 12:38