Should I learn AMD-specific APIs?

I am currently studying APIs related to Intel parallelization libraries such as TBB, MKL, and IPP. I was wondering, however, whether AMD should be considered part of the puzzle. Or is it just a waste of time? (I must admit, I have no idea about supporting the AMD library - in fact, I would be grateful for any advice that you may have.)

Just to clarify, the reason I am following the Intel path is because: 1) the APIs are very nice; and 2) Intel seems to support tool support as much as API support. (Once again, I have no clue how AMD works in this department.)

+5
source share
3 answers

MKL and IPP libraries will run (almost) also on AMD machines. I guess TBB will also work fine on AMD boxes. If I had to offer a technology that would be useful and useful for both, it would be an exploration of the OpenMP libraries. The Intel compiler with OpenMP extensions is amazingly fast and works with AMD chips.

+9
source

It is only worth it if you are interested in creating something like video games, operating systems, database servers or virtualization software. In other words: if you have a segment in which you care enough about performance to spend time on this (and do it right) in assembler. The same is true for Intel.

+7
source

Intel- , , AMD. ( ) , , . - . ( , .)

. , , Intel . , , Intel .

+3
source

All Articles