New Dolphin CPU Benchmark - NO GAME REQUIRED

[DJBarry004]

(01-26-2014, 01:28 AM)Nintendo Maniac 64 Wrote: Direct3D 9 is no longer included with new Dolphin revisions for a verity of reasons.

D3D9 was still included in 4.0-stable... And I do know the reasons.

[Nintendo Maniac 64]

And 4.0.x is pretty much it for Direct3D 9 going forward. The revision this benchmark uses is newer than 4.0.x and therefore doesn't even include it.


In other news, I have yet again become confused by the way the results chart is being updated...

The results for the i5-4570 from post #120 have been added to the chart, but several results that were posted before said i5, such as a Q8300 from post #118, are not listed on the results chart.

I'm going to go out on a limb as say delroth is doing this intentionally just to troll me.

[64DD]

CPU: i5-2500k @ 3.3GHz (OC: 4.3Ghz)
OS: Windows 8.1
Time: 10 minutes, 13 seconds
Screenshot: http://gyazo.com/15b89e93727ce972162694742fb4d315

[synce]

CPU: Intel i7 2760QM @ 2.4GHz (3.2GHz turbo boost w/all cores active)
OS: Windows 7 x64
Time: 14 minutes, 6 seconds
Screenshot: http://i.imgur.com/EWsnknd.png

I used throttle stop to ensure turbo boost never turned off during the bench

[Mush Man]

Could you possibly provide an x86 build to perform the benchmark on? I'd like to test on my other PC.

CPU: Intel Core i7 740QM @ 1.73 GHz (no OC), 2.93 GHz using Turbo Boost - Clarksfield series
OS: Windows 7 x64
Time: 26:01
more details

[BONKERS]

I retested my 4Ghz overclock again without doing anything else on the PC and my result improved by 50 seconds.

CPU: Core i7 950 OC @4Ghz
OS: Windows 7 Ultimate x64
Time: 11 minutes, 16 seconds
Screenshot: http://i5.minus.com/iN8tmNlqTc0gP.png



I also tested my CPU stock (3.06Ghz, 3.2Ghz Turbo)and it took ~14 minutes or so. I hit my screenshot key but for whatever reason the screenshot didn't save. So I rebooted into my OC and went to look for it and sighed when it wasn't there.

[admin89]

@Mush Man . Your CPU is not running at 2.93GHz , not even close ( I suppect 2.4GHz or even less)
2.93GHz is turbo boost with only 1 core active
http://www.notebookcheck.net/Intel-Core-i7-740QM-Notebook-Processor.31146.0.html
If your CPU run at 2.93GHz , your result should be the same as mine . However , my CPU is much faster , you see
Read the performance guide and try again :
https://forums.dolphin-emu.org/Thread-unofficial-laptop-performance-guide
Also , please use the correct software such as realtemp to see your CPU clock speed when running Dolphin , you can use throttlestop to set your CPU clock speed to a certain clock (as long as your laptop is not overheating)

[Mush Man]

(01-26-2014, 11:48 AM)admin89 Wrote: @Mush Man . Your CPU is not running at 2.93GHz , not even close ( I suppect 2.4GHz or even less)
2.93GHz is turbo boost with only 1 core active
http://www.notebookcheck.net/Intel-Core-i7-740QM-Notebook-Processor.31146.0.html
If your CPU run at 2.93GHz , your result should be the same as mine . However , my CPU is much faster , you see
Read the performance guide and try again :
https://forums.dolphin-emu.org/Thread-unofficial-laptop-performance-guide
Also , please use the correct software such as realtemp to see your CPU clock speed when running Dolphin , you can use throttlestop to set your CPU clock speed to a certain clock (as long as your laptop is not overheating)

Sorry, I was just going by online documentation for my CPU. Is the result correct if you instead take my base clock speed? If not, what exactly should I do to produce a valid test?

[u_liekmudkipz]

(01-22-2014, 01:39 AM)Jhonn Wrote: CPU: Intel Core i5 3230M @ 2.6GHz (Turbo Boost: 3.2GHz)
OS: Windows 8.1
Time: 16 minutes, 36 seconds
Screenshot: http://i.imgur.com/rmwlJDF.png

My laptop uses the same processor like your post but your result is a little bit faster than mine. What's the GPU of that unit? And that processor reaches about 3,100MHz during the benchmark? The highest frequency that my laptop's CPU could get while running the benchmark is 3,070MHz and it's highest temp is 80C. I'm just wondering if that extra 100+MHz matters with Dolphin's performance.

Full specs:
CPU: i5-3230M @ 2.6GHz
GPU: HD7670M 1GB
RAM: 8GB DDR3
OS: Win 8.1

[NaturalViolence]

Since I haven't gotten a response from delroth I went ahead and did it myself to see what the numbers would be: https://docs.google.com/file/d/0B9V1dG6cTJDnVWVDaDhiaGdQYlU/edit

Interestingly the IPC figures match the ratios observed in the WW benchmark perfectly. And I mean perfectly. Nehalem, sandy bridge, and ivy bridge (assuming quad core) differ by only 5%. Haswell is 30% higher than ivy bridge. Pretty much all architectures take a 5-10% loss when they have a smaller or absent L2/L3. Piledriver with L3 is equalvalent to penryn. Nehalem is around 15% higher than both. Without the L3 cache piledriver is equivalent to K10. Bulldozer is slower than that. Followed by K8 and finally jaguar.