Forum Discussion

Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
8 years ago

NDRnage Kernels Global Memory Write Pattern

I came up with some basic questions, which I would like to discuss: 1) I still don't understand how SIMD is being implemented in FPGA. In GPU, workitems being assigned to SIMDs are being execut...
Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
8 years ago

Yes I'm not reading anything from memory, since I want to role out memory effect as much as possible. But I don't understand why tempout is the same for all! it's a private variable in the kernel. Beside, none of the operations perform on constant values. As a result compiler cannot optimize them out. I have checked the verilog code to make sure all fmas are out there.