The Legend of Zelda: The Wind waker CPU Benchmark

[NaturalViolence]

If he did manage to get the thing to run WW at 6.5GHz 1.6v without deliding it and using phase change/LN2 cooling then he is literally the only person on the planet that can do it. That would be nothing short of a complete miracle. It would basically indicate that he has an ivy bridge chip that is somehow orders of magnitude more power efficient than every other known ivy bridge cpu in existence. Which is why I'm skeptical.

The 926% gamespeed reading is obviously wrong. That's 3 times the speed that 5GHz sandy/ivy bridge users are getting. Something is very wrong with his dolphin build.

@RDilus

Are you absolutely sure that you used the build from page 1 of this thread, didn't change any of the settings, and waiting 5+ seconds for the fps counter to stabilize?

[noah]

I'm going to upload my results tomorrow. When I did it I noticed that unlike when I normally play the game, it was running overly fast. So I put the graphics settings to 3xIR and it was still over 100%. 3.5xIR made it at 100%. I found this very strange for my integrated GPU. I usually can play at 1.5xIR at full speed. 2xIR is 90%. Why is this? And is it safe to run at 3.5xIR or would my laptop overheat?

[NaturalViolence]

noah Wrote:I usually can play at 1.5xIR at full speed. 2xIR is 90%

What do you mean by this?

Do you mean that you can normally play WW at these settings/speeds except in this benchmark?
Do you mean that you can normally play other games at these settings/speeds?
Or do you mean something else?

noah Wrote:And is it safe to run at 3.5xIR or would my laptop overheat?

I don't see why it would. Your cooling system should be capable of delivering the amount of cooling needed by the hardware. Manufacturers that don't do this will inevitably acquire a bad reputation and either go out of business or end up catering to a niche market so that users don't have a choice in the matter.

Not to mention Intel IGPs generate hardly any heat at all.

[noah]

I mean that on Dolphin 3.5 I can only get up to 2xIR before it gets annoying. But on the settings on Dolphin 3.0 which I got from this benchmark, I could run the game at 4xIR with full speed gameplay. I never changed the settings on either of these versions except for the IR.

Why is this? I thought Intel HD Graphics were bad, but 4xIR? Could it just be that Dolphin 3.0 is that much less demanding than Dolphin 3.5?

[Shonumi]

3.0 is a lot less demanding than 3.5, both on the CPU and GPU. You'll get better speed (albeit with less inaccurate emulation) in older revisions, generally speaking.

[NaturalViolence]

I think it has more to do with the fact that this benchmark is set to use efb copy to texture while 3.5 is set to use efb copy to ram via the game .ini. Efb copy to ram is MUCH slower on this game since this game makes heavy use of efb ram copies for post-processing effects. You need a much faster GPU to compensate for the video thread stall time introduced by the transferring the ram copies between main memory and video memory (as well as encoding/decoding them).

@noah

Set efb copy to ram in 3.0 and check to see if I'm right.

[Shonumi]

Just checked SS's builds that he provided. The GZLE01.ini file is for NTSC-U versions of this game correct? GZLP01.ini should be PAL, since that's what I get whenever I click "Edit Config". Neither the x64 or x86 builds have the necessary lines to disable EFB Copies to RAM for NTSC-U versions, but they're present for the PAL and NTSC-J versions. Did I miss something or am I just not looking thoroughly enough?

[noah]

Just got home. Alright I'll try it at around 7:20 (2 hours if your time is different).

[NaturalViolence]

Just say 19:20 [Insert timezone here]. Or 7:20 P.M. [Insert Timezone here].

[Starscream]

Garteal: I guess you can give me access to the spreadsheet if you like, I'd like to add my updated Llano results. Or, if you'd rather not, just update the Llano results.