Dolphin, the GameCube and Wii emulator - Forums

Hey Guys, thank you for sharing your data.

On a buyer perspective i am wondering how i could use the results for buying decision?

obv the lower the time the faster the cpu, but do i really need that 7 min benchmark time? Or is that an overkill and a CPU that does job in 15min is also fast enough for playing GC games like Soul Calibur 2?

Just submitted my 7980xe run, got 4m 33s. Not too shabby.
Running HT OFF @ 4.7ghz 1.22V all core on a X299 Omega no delid.
https://drive.google.com/open?id=1WtzCIT...lJhSPT2NYc

I've just submitted two.

How long would this test take on a real Wii?

To the person that submitted results from the ARM CPU in the Surface X, I've had to delete the result because it was not benchmarked on Dolphin 5.0 stable (it instead used a recent development version)

Yes this benchmark uses an old version of Dolphin, but we need to use a single consistent version in order for results to actually be comparable between different CPUs - this is meant to be a benchmark of the relative capabilities of how well various CPUs can handle emulation workloads in general, not a benchmark of how well Dolphin itself can run on a given CPU (since actual absolute performance tends to be slower).

And since newer versions of Dolphin perform better, it would give the impression of your CPU's emulation capabilities are better compared to other CPUs than they really are.

(11-06-2019, 02:56 AM)Intruso Wrote: [ -> ]How long would this test take on a real Wii?

According to Anandtech, around 1050 seconds. However, this is a bit misleading since this homebrew benchmark application tends to perform considerably better than actual games do in Dolphin. So even if your CPU gets a score that's 3 times faster than a real Wii console, you're unlikely to actually get such performance when running retail GameCube and Wii games in Dolphin.

Hmm, this seems wrong (AMD Ryzen 7 5800x), I'll test more later.



Edit: Looks like someone else got similar results with a 5600X so I guess that's just how it goes.

Hi, figured I'd register and upload a result for an AMD 5800X. Note that this is in ECO mode, because it runs at a flat 90C on full turbo. Until I sort out the cooling problem, I'm backing it off a bit, so this is more or less the result you'd get with a hypothetical 5700X.

This is with very good, DDR4-3600, 16-16-16-36-1T RAM:



edit: brought the image inline. Sorry, I'd have done that before, but the post wasn't visible until recently.

second edit: slight wording change, added DDR spec.

If by some lucky chance the submitter "Dolphinuser57" is reading this, please consider editing your submission and removing the unit at the end of your time (so rather than "322s" instead have it just say "322" without quotes) as it will greatly help the sorting of the chart.


And to the person (airidosas252) that submitted results from an ARM-powered Android device, I've had to delete the result because it was not benchmarked on Dolphin 5.0 stable since AFAICT a stable 5.0 version of Dolphin was never made available for Android.

As I said above, I realize that this benchmark uses an old version of Dolphin, but we need to use a single consistent version in order for results to actually be comparable between different CPUs - this is meant to be a benchmark of the relative capabilities of how well various CPUs can handle emulation workloads in general, not a benchmark of how well Dolphin itself can run on a given CPU (since actual absolute performance tends to be slower).

And since newer versions of Dolphin perform better, it would give the impression of your CPU's emulation capabilities are better compared to other CPUs than they really are.

(01-27-2021, 07:52 AM)Nintendo Maniac 64 Wrote: [ -> ]And to the person (airidosas252) that submitted results from an ARM-powered Android device, I've had to delete the result because it was not benchmarked on Dolphin 5.0 stable since AFAICT a stable 5.0 version of Dolphin was never made available for Android.

If you download 5.0-9, that version is functionally identical to 5.0. But I'm not sure if that's what they did.

(01-27-2021, 08:03 AM)JosJuice Wrote: [ -> ]If you download 5.0-9, that version is functionally identical to 5.0. But I'm not sure if that's what they did.

Good to know; I'll add mention of that to the first post. EDIT: I've simplified the instructions under "geeks that know what they're doing" by instead simply linking to 5.0-9.


BTW, am I allowed to ask if Dolphin has permanently transitioned to a beta-only rolling-release model, meaning v5.0 is effectively the last ever stable release?

Basically I've been wondering if should start thinking about an alternative solution to this benchmark situation that can still work well with a beta version, or whether to keep holding out for a newer stable release (I mean I've waited 4 years already, so I'm perfectly fine waiting longer - it's more about whether I will literally, not figuratively, be waiting forever if there will indeed be no more stable releasing of Dolphin)

...but if I do switch to using a newer beta, one starts having the question of "what beta version should I use?" I mean, places like Anandtech actually use this benchmark so I don't want to choose a random version (maybe I could base it around future hardware releases, like simply choosing the latest beta right before a very major new platform launches such as DDR5 desktop CPUs?).

(01-27-2021, 08:09 AM)Nintendo Maniac 64 Wrote: [ -> ]BTW, am I allowed to ask if Dolphin has permanently transitioned to a beta-only rolling-release model, meaning v5.0 is effectively the last ever stable release?

The plan is still to release a 6.0 at some point, but it's not happening in the near future (I wouldn't expect it to happen this year, at least).