News for hackers. Stuff that matters.
Re:Matrix Multiplication Inches Closer To Mythic Goal by reportingsjr on Saturday December 18, 2021 @12:26PM (#29606365)
If you look at operations per cycle, in this example for Intel Haswell processors, you'll find that an equivalent add and multiple are quite a bit different.

An add/subtract from a 32 bit register to register is 0.25 cycles per operation.

A multiply from 32 bit register to register is 2 cycles per operation. So a 8x speed difference, not counting for latency, etc.

See page 230 and 231 of this paper: https://www.agner.org/optimize/instruction_tables.pdf

Re:Matrix Multiplication Inches Closer To Mythic Goal by gnufx on Saturday December 18, 2021 @01:12PM (#29606876)
For what it's worth, what counts for normal implementations is the number of multiply-add operations per cycle, typically two on current hardware with FMA.