Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
New Cpu Benchmark test Using Pcsx2 Bios
#1
The goal of this Benchmark is to create a universal CPU benchmark that doesn't rely on any particular game, but gives reliable info on the performance of the CPU
In order for the results to be meaningful it is required that certain standards are met, they are as followed

1. The Same version of Pcsx2 will be usedPcsx2 0.9.8
2. No speed Hack`s will be allowed(this includes the Mufti Thread hack)
3. Only Software Mode will be used in Video plugin
4. OC CPU must be stated if used

The Test itself is Simple and Only requires running your Ps2 bios with Frame Limiting turned off(do not run a disk or image )
The Point to take the FPS reading is the Browser/System Configuration Page
The Maximum Resulting FPS is then your CPU`s Performance and can be presented as Follows
1.CPU
2.FPS
example
Cpu i7 2670OM @2.2Ghz
Fps 144
I hope these instructions are presented in an easy to understand light



Reserved
ROG Asus G74S laptop
Intel core i7 2670QM @2.2(3.1ghz turbo) Nvidia 3gig GTX560m 8 gigs DDr5 Blue-Ray Burner windows7 64 bit
[Image: r1gyld.jpg]
Reply

Sponsored links

#2
My benchmark on AMD Phenom II X4 @3.0Ghz

35 Average (r4600) / 75 FPS average (r5100)

Default Emulator Settings
800x600 window size
Software Rendering

Frankly, the SVN is a winner all around, since when 0.9.9 will officially be coming out, all those benchmarks will be a waste given the huge performance differences.
Reply
#3
(02-23-2012, 05:46 PM)Coornio Wrote: ... all those benchmarks will be a waste given the huge performance differences.

You don't understand.
A benchmark isn't just to see the highest FPS achievable from the given scene (if even at all), nor to compare revisions of PCSX2. (Although, it's good to see it higher in comparison there. Smile).

Benchmarks are usually for a direct, and accurate (as possible) comparison between different hardware. Even if the newer revision is xx% faster, it's not like the number scored reflects anything except how it compares. (Unless you like staring at the BIOS menu. ;P)
Reply
#4
(02-24-2012, 06:46 AM)Rezard Wrote: You don't understand.
A benchmark isn't just to see the highest FPS achievable from the given scene (if even at all), nor to compare revisions of PCSX2. (Although, it's good to see it higher in comparison there. Smile).

Benchmarks are usually for a direct, and accurate (as possible) comparison between different hardware. Even if the newer revision is xx% faster, it's not like the number scored reflects anything except how it compares. (Unless you like staring at the BIOS menu. ;P)

On the contrary, i like staring at high FPS rates Laugh
Reply
#5
My concern would be the different bios & regions out there they could operate differently
Intel Core I7 930@3.5GHz(TB, HT), 6GB DDR3@1600MHz, Geforce 970 G1
Reply
#6
By using software rendering in GSdx what you end up measuring is GSdx and not PCSX2 or the CPU, as it's been shown already a different revision of GSdx made all the difference in the test, also software rendering is not something most people will use what makes the bottleneck here is not quite what you'll find in most games, and you would not be using software mode for most games either way. This is not anything good as far as a benchmark goes.
Core i5 3570k -- Geforce GTX 670  --  Windows 7 x64
Reply
#7
(02-25-2012, 11:04 PM)Shadow Lady Wrote: By using software rendering in GSdx what you end up measuring is GSdx and not PCSX2 or the CPU, as it's been shown already a different revision of GSdx made all the difference in the test, also software rendering is not something most people will use what makes the bottleneck here is not quite what you'll find in most games, and you would not be using software mode for most games either way. This is not anything good as far as a benchmark goes.
I would have thought if you eliminate the GPU then all you have left is the CPU, and since software mode doesn't use the GPU it must then use the CPU
And i was also stating use of only one version of Pcsx2 , any who its not been the most successful thread Tongue


ROG Asus G74S laptop
Intel core i7 2670QM @2.2(3.1ghz turbo) Nvidia 3gig GTX560m 8 gigs DDr5 Blue-Ray Burner windows7 64 bit
[Image: r1gyld.jpg]
Reply
#8
you're putting the load on GSDX instead of the main core by using software rendering.
Reply
#9
(02-23-2012, 05:46 PM)Coornio Wrote: My benchmark on AMD Phenom II X4 @3.0Ghz

35 Average (r4600) / 75 FPS average (r5100)

Default Emulator Settings
800x600 window size
Software Rendering

Frankly, the SVN is a winner all around, since when 0.9.9 will officially be coming out, all those benchmarks will be a waste given the huge performance differences.

Sad how a 900mhz increase with the same CPU only nets me 5 extra FPS on the same SVN haha..

Anyway. Phenom II x4 965BE @ 3.9ghz
SVN 5100 80FPS (114 was highest, lowest was 80, also what it was stable at)
Intel Core i5-4690k @4.5ghz
Samsung 16GB DDR3 @2000mhz
GeForce GTX 680 4GB
Windows 7 x64
Reply
#10
(03-10-2012, 07:26 PM)hellbringer616 Wrote: Sad how a 900mhz increase with the same CPU only nets me 5 extra FPS on the same SVN haha..

Anyway. Phenom II x4 965BE @ 3.9ghz
SVN 5100 80FPS (114 was highest, lowest was 80, also what it was stable at)

I wonder if mine would procure any more frames if overclocked. Despite not being a black box release, it was completely unlocked with my previous motherboard, which would imply that either the core was unlocked for some reason or that the motherboard was somehow unique. In my current one, i cannot change a single thing regarding the CPU without making the pc unbootable. Goddamn DDR3 upgrade killed me..
Reply




Users browsing this thread: 1 Guest(s)