Forum Discussion

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

Short vs Int vs Floating Point usage in Kernels

I'm just wondering if anyone else has successfully used smaller fixed point representations for data in their kernels compared to floating point. By that I mean I'm playing around with changing my a...
Altera_Forum's avatar
Altera_Forum
Icon for Honored Contributor rankHonored Contributor
11 years ago

--- Quote Start ---

Why would there be wasted memory if you're packing 4 char values into an integer? Or do you mean wasted memory in terms of logic elements used to convert (mask) from the 32 bits down to the chars and back.

I can't exactly explain why the DSP increased for the dot product

--- Quote End ---

I was just referring to storing chars as integers in memory, i.e. using only 8-bits out of 32.

Integer addition does not use DSP blocks, but multiplication does, hence, the difference in DSP usage.