Dolphin, the GameCube and Wii emulator - Forums

Here are my Results for 4.8ghz

test1 97
test2 79
test3 120

saw some decent improvements of fps going up by 300mhz. 5+ fps on each one.

going try 5ghz now












And here is my 5ghz Scores.


test1 103
test2 84
test3 129






these 3 posts were unmodified ram only CPU overclock. Will test if ram oc will help in dolphin later.

My results for my HTPC
A10-5800K (stock 3.8Ghz with turbo on)
8Gig 1600Mhz
Windows 8 64BitNTSC version


Test 1 47
Test 2 41
Test 3 62

I'll just use dropbox links to the pics
https://dl.dropbox.com/u/12166653/Dolphin%201.png
https://dl.dropbox.com/u/12166653/Dolphin%202.png
https://dl.dropbox.com/u/12166653/Dolphin%203.png

(02-08-2013, 11:57 PM)Runadumb Wrote: [ -> ]My results for my HTPC
A10-5800K (stock 3.8Ghz with turbo on)
8Gig 1600Mhz
Windows 8 64Bit

Test 1 47
Test 2 41
Test 3 62

At first I was like "there is NO WAY another Trinity APU user is getting such slow speeds, heck my Brisbane is faster!". Then I realized you were quoting FPS, not VPS.

Now we just need an FX-8350 owner to run this bench and maybe we'll actually get an idea how Piledriver does. I've been in contact with an FX-8350 owner that is a Dolphin user, but he said he'd do the bench later...who knows how much later "later" is.

(02-08-2013, 11:57 PM)Runadumb Wrote: [ -> ]Test 1 47
Test 2 41
Test 3 62

Interesting how for you Test 1 is faster than Test 2 while for me it's the other way around. I assume you're using the integrated graphics card and not a dedicated one?!


In percentage your numbers are, 68%, 28% and 55% better than mine. Which is a lot more than the usual 5-10% I read in other benchmarks of the two APUs. So clearly something is wrong with my system. I just wonder what it might be. I have all Mobo Settings on Auto with the latest Bios and it's not like there's a Turbo Button on the front of my case which I forgot to press :-)

CPU-Z tells me the Cores are running at the desired Speed, RAM has the correct voltage and speed as well. Dunno what else there is to check. I have no background tasks with high CPU usage or anything like that running, which of course was the first thing I checked. Nothing is doing excessive I/O on the disks... really don't know where else to look.

Any Ideas?

(02-10-2013, 01:36 AM)Luu Wrote: [ -> ]Interesting how for you Test 1 is faster than Test 2 while for me it's the other way around.

Thats the case for all but just a few people on this benchmark. Your test 2 is the only one that seems right. It has been a while since I read your other thread so I can't remember what all you tried. Have you tried a fresh install of windows? Did you make sure Turbocore wasn't disabled in your bios? You can see in two of runadumb's pics that his cpu is running at ~4ghz Thats a 600mhz difference if turbo isn't working for you.

(02-08-2013, 11:57 PM)Runadumb Wrote: [ -> ]My results for my HTPC
A10-5800K (stock 3.8Ghz with turbo on)
8Gig 1600Mhz
Windows 8 64Bit

Test 1 47
Test 2 41
Test 3 62

I put a new cooler on my 1100t and my new scores are:
Test 1 47
Test 2 39
Test 3 63

Still at stock 3.3ghz turbo 3.7. Ill get around to posting pics and full specs later. Ill also retest with turbocore disabled to see just how much it is helping. You can see that even with a 500mhz advantage on the trinity they virtually perform the same. So luu's 5700 should perform about on par with a 2.9-3ghz phenom ii or slightly behind the b50 @ 3.1ghz I posted a while back. The B50 got 42 35 and 57 so just as a guess the 5700 should be getting somewhere around 40 33 and 55 for the tests.

OS: Windows 7 home premium x64
CPU: Phenom II x6 1100t (Stock)
GPU: Radeon HD 5570 ddr3 version
RAM: 8GB Crucial Ballistix 1600
Dolphin: 64bit From this thread only options changed where the ones Starscream changed and auto adjust window to make pics not as big.
GAME: NTSC

Save 1


Save 2


Save 3



Disabling Turbo dropped the second save 2 fps dropped the third save 1 fps and made the first save more unstable though it was still 47 at the same spot. It's possible that the jump from ddr2 800 to ddr3 1600 helped boost performance as well as the new cooler. I noticed the stock cooler, from my initial post, was very convex on the bottom. So much so that you could tell just by looking at it without putting something flat against it.

Now the same setup with cpu overclocked to 3.8 ghz and turbocore disabled.
Save 1


Save 2


Save 3

(02-10-2013, 01:36 AM)Luu Wrote: [ -> ]Interesting how for you Test 1 is faster than Test 2 while for me it's the other way around. I assume you're using the integrated graphics card and not a dedicated one?!

Yes I'm just using the onboard graphics. I've no idea why you are getting such poor performance but I will say my build wasn't smooth sailing either. I had to RMA the SSD, change memory (mine was foiling the heatsink) and do a full windows reinstall as it was crazy unstable. All is fine now though.

Good to know that the trinity (piledriver) results are exactly what I expected based on similar benchmarks. Ivy bridge is beating it clock for clock by about 55-65%. Sandy bridge by about 40-50%. Phenom II is beating it in IPC by exactly 12% in all three tests. Vishera (piledriver) has roughly the same IPC as phenom II in similar tests and 10% higher IPC than trinity so this is right on the dot. We're still missing quad core nehalem, core 2, zambezi (bulldozer), vishera (piledriver), athlon II (10H), phenom (10H), atom (bonnel), E series (bobcat), and C series (bobcat). Unfortunately atom is out of the question since the IGP is too shitty to keep up even at 1x.

Runadumb can you retest with the cpu underclocked to 3.4GHz and turbocore off?

There are some Atom devices with Nvida Ion, surely they'd run the benchmark just fine.

Also I have a friend with an E-350 "notbook", maybe the next time he stops over I can run the benchmark. He also owns an FX-8150 but I rarely if ever go to his place.

And you know, you could always run this benchmark off a flash drive on in-store demo PCs. (I did that on a Sandy Bridge laptop once to see if an HD3000 could manage good performance on even a 2D game in Dolphin - it could not).

I wouldn't mind seeing Xalphenos's Phenom II x6 1100t running at 2.4GHz to compare with my CPU.

Quote:There are some Atom devices with Nvida Ion, surely they'd run the benchmark just fin

Yes that would work.

Quote:Also I have a friend with an E-350 "notbook", maybe the next time he stops over I can run the benchmark. He also owns an FX-8150 but I rarely if ever go to his place.

Give me his address and I will "encourage" him.

Quote:And you know, you could always run this benchmark off a flash drive on in-store demo PCs. Tongue

I think that would be a little weird.....

I can imagine a lot of people giving me funny looks as I run around best buy running GC/Wii games on all of their PCs.