Console emulator benchmarks at AT
- Thread starter kemosabe
- Start date
It's an interesting idea, but the fact that the Pete in question only tested on a 9700 may limit the usefulness of the GF numbers. OTOH, it is an OGL driver, and nV has been touted as having superior OGL performance.... 
I can't believe they have a GC emu already, even at 5fps. Man, that's fast.
I can't believe they have a GC emu already, even at 5fps. Man, that's fast.
It just seems there's a kinda irony that the 9800 PRO isn't better at the GC titles! Joe DeFuria
Legend
Coming soon to a Detonator near you: magical performance increaes in "emulators that anand uses" for benchmarks.
Either that, or the emulators themselves will enter the TWIMTBP program...
Either that, or the emulators themselves will enter the TWIMTBP program...
Joe DeFuria said:Coming soon to a Detonator near you: magical performance increaes in "emulators that anand uses" for benchmarks.
Either that, or the emulators themselves will enter the TWIMTBP program...
Boy, you're not jaded are you?
Anyway, I filed this under "I don't give a crap". Guess I'm right there with you.
Tommy McClain
StealthHawk
Regular
Seems like an interesting "for fun" thing to do, but it hardly seems worth the time for any serious analysis(like in actual reviews). This isn't production software, and most emulation stuff is pretty biased- developed and tested on specific hardware as already mentioned.
davefb said:well interesting, but personally
i hope nintendo (sony etc) come down on em like a ton of bricks .
just my 2c
-dave-
Why? They could very well have ripped their own store-bought TWW disc if they have a GC with a BBA, which is perfectly legal. Or, they could even have run Dolphin USING a Networked GC with the game in it, to run the disc directly... which is unquestionably legal. They could also easily have run a legally purchased FFIX in a normal CD-ROM drive.
And emulators, in case you didn't realise, are -quite- legal.
Joe DeFuria
Legend
StealthHawk said:
Yes, but if Anand keeps on using these benchmarks (which he claims he will do), do you think nVidia will sit and watch their product get hammered?
They'll either pressure Anand to "change his mind" about the benchmarks, start sending "Free hardware" to the developers of the emulators, or *cough* optimize *cough* for the emulator.
StealthHawk
Regular
Joe DeFuria said:They'll either pressure Anand to "change his mind" about the benchmarks, start sending "Free hardware" to the developers of the emulators, or *cough* optimize *cough* for the emulator.
I would welcome true optimizing as opposed to "optimizing." What's to stop ATI from sending free hardware to the developers of emulators if NVIDIA does it?
Have we really reached a point where we've gotten cynical enough to think that NVIDIA would cheat in emulation? Which is probably pretty hard to do considering that emulation greatly depends on which plugins for the emulator you use, how you set up those plugins, and what games you try to emulate. Seems like a lot more trouble than its worth.
It's not like all is perfect in ATI's world. Their (much slower than NVIDIA) slow framebuffer read speed is teh sux! It still hasn't been fixed yet AFAIK although they have been notified of it(see Catalyst bug list @ R3D), maybe this will move it up on ATI's priority list!
P.S. Pete used to test on GeForce cards before he got his ATI card. Plus NVIDIA cards can use more of his plugin features(higher resolution) than ATI cards. It's my opinion that the emulation scene is more biased to NVIDIA, but I could be wrong. I know a while back more things were developed with NVIDIA in mind.
That was the case with any backroom gaphics coding scene - the proliferation of NVIDIA boards and their developer support meant that just happened. I think the Stanford guys doing the Dawn wrapper was actually an indication that ATI had broken in that scene a little.
Joe DeFuria
Legend
StealthHawk said:
You missed my points entirely.
My point are:
1) ATI used to (and still does) get flack for drivers, when the "issue" at hand is primarily one of development platform. nVidia has been the platform of choice through the GF4 era. This is chanigng with the R300/NV3x.
2) It's obvious that nVidia themselves hasn't given 2 sh*ts about the emulator...except now that it's being used as a benchmark at a "popular" site, I expect this to change...and rather quickly.
StealthHawk
Regular
Joe DeFuria said:
Yes, this is something that is changing. It's certainly not a revolution that is over, nor may it be one that finishes with the tides turned. We're (probably) going to see this shift more in upcoming DX9 games. I hesitate to so easily say that console emulation has already made as much of a shift as PC gaming has made and will continue to make. My whole point was that it hasn't, and that emulation is NOT biased towards ATI. It's probably more equal or a little biased still towards NVIDIA, after all, most of these emulators and plugins were developed when NVIDIA was still king. Certainly something as old as ePSXe falls under this category. The same thing might not be said for things like PS2, XBox, and GC.
And I, as well as all other people who use emulators, will welcome legitimate optimization. From NVIDIA or ATI. Not sure where NVIDIA would cut quality, if that is what you are suggesting...PSX games already have crummy graphics and don't even use bilinear filtering unless you force it on, which causes some graphical glitches.
Tagrineth said:
I can't speak for N64, as I don't know how they do it, but for PSX, there are workarounds in place. But wouldn't it be nice if there didn't have to be? Also, I see drops when entering battles in games like FF7.
Joe DeFuria
Legend
StealthHawk said:
Actually, my point is not really along the lines of "what will nvidia do to kill quality to increase performance." (Though that may happen). It's more of "if nVidia's drivers were so robust as is the general thought, they wouldn't have to do any case by case optimization at all. And if you buy a game that's not popular enough to be a popular benchmark, you're getting "pot luck" with nvidia's drivers...particularly as DX9 titles get more and more popular.
StealthHawk said:
http://steelkiel.tripod.com/topswirl1.jpg
http://steelkiel.tripod.com/topswirl2.jpg
Copy and paste the URL's in a new browser (or tab)...
Looks like 59.96 fps to me! That's with AA and AF, and in a window...
FFIX does the same thing, at least for me.
Note, that this is with a version of Pete's um... I don't remember which API now, that was released BEFORE R300.
Edit: Added: And this is on a not-yet-overclocked Radeon 9500 Pro. (it's OC'd now, but wasn't at the time)
Have I said this is the most incredibly useful article by Anand by far?
Say, should I start talking about how much a video card contributes to the overall heat in my case? Or why the quality of popcorn is affected by the quality of a microwave oven? Or perhaps why one of the really important reasons I buy a sedan is based on how close it is to a Ferrari in in every consideration?
Say, should I start talking about how much a video card contributes to the overall heat in my case? Or why the quality of popcorn is affected by the quality of a microwave oven? Or perhaps why one of the really important reasons I buy a sedan is based on how close it is to a Ferrari in in every consideration?
