|
test the performance in MAME - benchmarks explained
#362046 - 12/31/16 06:44 PM
|
|
|
Hello,
I'm trying to find a detailed tutorial on how to know how well my system runs games.
I think that what I'm looking for is 'benchmarks in mame', but I don't know how to run/use them in mameui or mame.
If I have, let's say, Ridge Racer, I want to know:
- how to run mameui or mame to tell it to 'measure' the performance
- how can I get the results and understand them (fps, percent, and their meaning).
- How can I compare and submit my results with other systems.
The thing is that I have an old PC (Intel Pentium D 2.4 Ghz, 2Gb RAM, Nvidia GForce 705, Windows 8 x64) and I'm planning to get another one with enough power to run games on mame (and many other things of course).
Thanks in advance !
Roger
|
|
|
John IV |
IV/Play, MAME, MAMEUI
|
|
|
Reged: 09/22/03
|
Posts: 1981
|
Loc: Washington, USA
|
|
Send PM
|
|
|
Re: test the performance in MAME - benchmarks explained
[Re: rogerts]
#362074 - 12/31/16 11:11 PM
|
|
|
Easiest is to use the baseline version of Mame from the commandline.
mame ridgerac -bench 90
The switch is -bench, argument is time in emulated seconds. This tests the CPU basically w/ no sound or graphics and is useful for comparing different intel chips (or AMD which are historically worse)
Results are in percentages, 100% is running full out.
See http://www.mameui.info/Bench.htm
Note some of those games on the page need to be set up before running.
|
|
|
|
Re: test the performance in MAME - benchmarks explained
[Re: John IV]
#362081 - 01/01/17 12:47 AM
|
|
|
> Easiest is to use the baseline version of Mame from the commandline. > > mame ridgerac -bench 90 > > The switch is -bench, argument is time in emulated seconds. This tests the CPU > basically w/ no sound or graphics and is useful for comparing different intel chips > (or AMD which are historically worse) > > Results are in percentages, 100% is running full out. > > See http://www.mameui.info/Bench.htm > > Note some of those games on the page need to be set up before running.
How different would those benchmarks be if you test them with sound and graphics? Is there any specific methodology for doing it that way? There's obviously no -benchsoundgraphics commandline switch. I'm merely curious.
|
"Note to Noobs:
We are glad to help you but simply posting that something does not work is not going to lead to you getting help. The more information you can supply defining your problem, the less likely it will be that you will get smart-alec replies.
C.D.~"
|
|
|
Re: test the performance in MAME - benchmarks explained
[Re: Master O]
#362084 - 01/01/17 02:01 AM
|
|
|
> How different would those benchmarks be if you test them with sound and graphics? Is > there any specific methodology for doing it that way? There's obviously no > -benchsoundgraphics commandline switch. I'm merely curious.
If anything, it's a bit of apocrypha from MAME's history.
Back in the day, running with -nosound caused MAME to specifically exclude certain driver hardware when running the game. More recently, the decision was made for MAME to always emulate the sound hardware - mostly to reduce bug reports from people running with -nosound.
So, nowadays, there shouldn't be much (if any) difference between running with -sound none and running with normal MAME. In the past, though, it might have gained you some frames per second, at the cost of potentially fucking up the emulation of the game itself.
|
|
|
Tafoid![Administrator Administrator](//www.mameworld.info/ubbthreads/images/adm.gif) |
I keep on testing.. testing.. testing... into the future!
|
|
|
Reged: 04/19/06
|
Posts: 3138
|
Loc: USA
|
|
Send PM
|
|
|
Re: test the performance in MAME - benchmarks explained
[Re: Master O]
#362085 - 01/01/17 02:04 AM
|
|
|
> > How different would those benchmarks be if you test them with sound and graphics? Is > there any specific methodology for doing it that way? There's obviously no > -benchsoundgraphics commandline switch. I'm merely curious.
Sure there is.. it is called -str (-seconds_to_run). -bench is used because MAME isn't about sound and graphics cards, but the lowest common thread in all cases where emulation is concerned being how fast can it crank out emulated opcodes.
|
|
|
|
Re: test the performance in MAME - benchmarks explained
[Re: Tafoid]
#362086 - 01/01/17 02:56 AM
|
|
|
> > > > How different would those benchmarks be if you test them with sound and graphics? > Is > > there any specific methodology for doing it that way? There's obviously no > > -benchsoundgraphics commandline switch. I'm merely curious. > > Sure there is.. it is called -str (-seconds_to_run). > -bench is used because MAME isn't about sound and graphics cards, but the lowest > common thread in all cases where emulation is concerned being how fast can it crank > out emulated opcodes.
My mistake. Thanks for clearing that up.
|
"Note to Noobs:
We are glad to help you but simply posting that something does not work is not going to lead to you getting help. The more information you can supply defining your problem, the less likely it will be that you will get smart-alec replies.
C.D.~"
|
|
|
Re: test the performance in MAME - benchmarks explained
[Re: MooglyGuy]
#362087 - 01/01/17 03:01 AM
|
|
|
> Back in the day, running with -nosound caused MAME to specifically exclude certain > driver hardware when running the game. More recently, the decision was made for MAME > to always emulate the sound hardware - mostly to reduce bug reports from people > running with -nosound.
Anyone know when this was?
Edit: Thanks.
|
MAME Artwork / Artwork Previews
|
|
Tafoid![Administrator Administrator](//www.mameworld.info/ubbthreads/images/adm.gif) |
I keep on testing.. testing.. testing... into the future!
|
|
|
Reged: 04/19/06
|
Posts: 3138
|
Loc: USA
|
|
Send PM
|
|
|
Re: test the performance in MAME - benchmarks explained
[Re: hydef]
#362088 - 01/01/17 03:38 AM
|
|
|
> > Back in the day, running with -nosound caused MAME to specifically exclude certain > > driver hardware when running the game. More recently, the decision was made for > MAME > > to always emulate the sound hardware - mostly to reduce bug reports from people > > running with -nosound. > > Anyone know when this was?
The whatsnew.txt shows it officially happened in MAME 0.102, not exactly recently, but the idea of disabling cpu cores and not loading sound roms never was a great idea.
|
|
|
|
|
I've attached my latest version of the bench batch file.
It has instructions at the beginning on exactly how to set up the games that require it.
I haven't done any benchmarking in a while, but I'm pretty sure it will work with the latest version of MAME as long as you have all the required ROMs.
|
GroovyMAME support forum on BYOAC
|
|
|
Re: test the performance in MAME - benchmarks explained
[Re: Tafoid]
#362105 - 01/01/17 01:20 PM
|
|
|
> > > Back in the day, running with -nosound caused MAME to specifically exclude > certain > > > driver hardware when running the game. More recently, the decision was made for > > MAME > > > to always emulate the sound hardware - mostly to reduce bug reports from people > > > running with -nosound. > > > > Anyone know when this was? > > The whatsnew.txt shows it officially happened in MAME 0.102, not exactly recently, > but the idea of disabling cpu cores and not loading sound roms never was a great > idea.
yeah disabling emulation of the actual hardware was probably one of the worst ideas we ever had, caused so many little issues that nobody ever quite fully kept track of.
|
|
|
R. Belmont |
Cuckoo for IGAvania
|
|
|
Reged: 09/21/03
|
Posts: 9717
|
Loc: ECV-197 The Orville
|
|
Send PM
|
|
|
Re: test the performance in MAME - benchmarks explained
[Re: Master O]
#362109 - 01/01/17 04:07 PM
|
|
|
> How different would those benchmarks be if you test them with sound and graphics? Is > there any specific methodology for doing it that way? There's obviously no > -benchsoundgraphics commandline switch. I'm merely curious.
Just to clarify: the sound and graphics are still emulated with -bench, they just aren't output. This removes your OS and audio/video drivers and gives a pure CPU benchmark.
|
|
|