lilypond-user
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: CPU stress tests for LilyPond


From: Paolo Prete
Subject: Re: CPU stress tests for LilyPond
Date: Fri, 3 Dec 2021 01:03:26 +0100



On Thu, Dec 2, 2021 at 9:21 PM David Wright <lilylis@lionunicorn.co.uk> wrote:
On Thu 02 Dec 2021 at 09:05:08 (+0100), Thomas Scharkowski wrote:
> On Thu, Dec 2, 2021 at 2:41 AM Paolo Prete wrote:

> > With my processor (Intel Celeron N3350) it took 96 seconds to compile MSDN.ly
> >
> > But my CPU is listed with score 287:

That too seems slow. What OS is it running?


Thanks for noting this.  I re-ran the tests, and I confirm that it took 96 secs on LilyPond 2.20.0-1 on Lubuntu (newest version). Same time without the X environment. However, with LilyPond 2.22.0-1 it took 55 seconds. This is a huge difference. Now, if I compare this last result with the Apple Mini of Jacques, I obtain 55 seconds versus 39 seconds. But the Gzip test shows 91 versus 40, then the Gzip test is not a reliable parameter too for our test.

 
> > https://openbenchmarking.org/vs/Processor/Apple%20M1,Intel%20Celeron%20J3455
> >
> > (note that it compares Apple M1 with Celeron J3455, which is NOT my CPU, but it should be very similar for single-core tests.)
> >
> > The result is 91 / 40 which is very similar to 96 seconds / 40 seconds (Jacque's test)  . I don't know if this is a coincidence and please, if you all have more data, share it.

I can't get results from that site because only the
fastest machine above matches: there's nothing comparable
even to the middle one, and the slowest probably raises
some eyebrows.

Yeah, there are not many machines for the Gzip test. But did you check for a similar CPU on the big list? 

https://openbenchmarking.org/test/pts/compress-gzip

Anyway, we have to fix some points before doing further tests:

1) which version of LilyPond should we use? I would propose 2.22.0
2) which test can we use, as a reference? Geekbench 5 has the disadvantage I said before, and Gzip doesn't seem to me the best one, give the discrepancy between 55/39 and 91/40

Cheers,
Paolo


 

Cheers,
David.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]