search

JulianKemmerer / PipelineC

A C-like hardware description language (HDL) adding high level synthesis(HLS)-like automatic pipelining as a language construct/compiler feature.
https://github.com/JulianKemmerer/PipelineC/wiki
GNU General Public License v3.0
607 stars 50 forks source link

Make tool optimize for fewer pipelining registers #64

Closed JulianKemmerer closed 2 weeks ago

JulianKemmerer commented 2 years ago

Right now the tool inserts registers to meet timing - with no regard to how many registers are being inserted.

Instead: manually find the smallest bitwidth wire ~in the area... always choose to slice fewer wires over more if possible with the same timing characteristics/delays ...

JulianKemmerer commented 2 years ago

https://google.github.io/xls/scheduling/#scheduling-to-minimize-pipeline-registers

JulianKemmerer commented 2 years ago

Who knows maybe https://github.com/JulianKemmerer/PipelineC/blob/master/src/DEVICE_MODELS.py will get to this some day?

JulianKemmerer commented 2 years ago

Closing https://github.com/JulianKemmerer/PipelineC/issues/65 definitely helps reduce register use if you are combining global variables and deep pipelines

suarezvictor commented 2 years ago

Does it lower resource usage? Any measutement?

JulianKemmerer commented 2 years ago

Yeah recent commit for #65 looks to save ~50 percent of flip flops on ECP5 ray tracer builds - we should do another attempt making it work

On Artix I measured up to like 87K flip-flops saved depending on how your count shift registers.

See info in the pipelinec discord 👍

suarezvictor commented 2 years ago

Excellent work

bartokon commented 2 years ago

What about keccak resource usage? I'm curious 😏

JulianKemmerer commented 2 years ago

@bartokon for keccak - the particulars of #65 I wouldnt expect to apply much IIRC about keccak 1) since it doesnt maintain a global state variable wired into the pipeline at various places or 2) doesnt have lots of 'unresolved compile time math'/placing doing math on constants that is mistaken for stuff needing LUTs, etc

There also could be some room to tune some of weird pipelining magic numbers ... I am experimenting with some things right now for the ray tracer

So ~maybe but probably not much improvement

JulianKemmerer commented 1 year ago

Similar issue is wanting to optimize for area

starts with first parsing area out of the reports from the tool...

Similar to #45