Bandwidth Place 2

thanx 05. The only thing I want to replace, not important now but sometime, is to get a better PS. I guess I was looking for a benchmark program that I could compare scores to is all. Not sure if that's even done anymore LOL.

No problem

Your right on the power supply though. You do not want a cheapo power supply destroying the other very expensive parts of your computer when it blows up on you.

Stick with high quality name brands like Corsair, Antec, Enermax, Thermaltake and so on.

As far as a benchmark program your really looking at something like Geekbench will test your CPU/RAM. Your card is already proven powerful and still considered "up to date".

Give geekbench a try. I'm gonna run it on my laptop and the rig in my signature and post up the results for fun.

Geekbench Score-5356
Section-------Description--------------------------Score
Integer-------Processor integer performance-------5806
Floating Point-Processor floating point performance-7028
Memory-------Memory performance-----------------3094
Stream-------Memory bandwidth performance------2457
Geekbench 2.3.1 Tryout for Windows x86 (32-bit)

Well if we're going to paste the results as text:

Geekbench Score-12870
Section-------Description--------------------------Score
Integer-------Processor integer performance-------13899
Floating Point-Processor floating point performance-16351
Memory-------Memory performance-----------------7397
Stream-------Memory bandwidth performance------8035
Geekbench 2.3.1 Tryout for Windows x86 (32-bit)

otherwise, http://browse.geekbench.ca/geekbench2/709904

Wow, more then double mine. nice

Geekbench Score-7536
Section-------Description--------------------------Score
Integer-------Processor integer performance-------7898
Floating Point-Processor floating point performance-9423
Memory-------Memory performance-----------------4785
Stream-------Memory bandwidth performance------5176
Geekbench 2.3.1 Tryout for Windows x86 (32-bit)

or

http://browse.geekbench.ca/geekbench2/711059

Mines just a tad bit higher then Directives. You should have no issues Dir

PB, eat one brother! That's an impressive score.

I was thinking, I have like 17 things in my system tray. will that effect my score?

Sure, but then again... I've got a bigger problem with the fact that Geekbench free is only 32bit. I liked Cinebench which is only CPU dependent. Directive, perhaps you'd like to contribute to the charts?
I'll need 32 and 64bit tests for both R10 and R11.5, but R11.5 takes priority.
http://www.mylilsite.net/images/cineben ... latest.png
http://www.mylilsite.net/images/cineben ... latest.png

If we want something that gives a score and is dependent on the CPU and RAM, I like IntelBurnTest. It contains linpack binaries for 32 and 64bit and the correct one is automatically chosen. I also like it because it's up to date with new instruction extensions added to modern CPUs, unlike traditional things like prime95. But... that's where you guys won't like it. IntelBurnTest uses AVX instructions that are new to my CPU generation and makes my score far beyond your reach.


http://www.mylilsite.net/images/sandybr ... 500mhz.png

If my memory serves, the score I got back before Windows 7 SP1 (SP1 added AVX support) was about 45GFlops.

I can tell. Just ran that program and got the following...



Not even close LOL

Didn't you get like 13GFlops earlier? What happened?

I'm gonna run that program tonight when i get home and post up the results just for shits!

Isn't 34 better then 13? Not sure what your asking.

How did you get 13 if you can get 34? Even if you got 34 at 4GHz and 13 at 2.8GHz... that math ain't right.

ok, like I said before I have no idea what these numbers mean. SHould I be looking for something specific?

I didn't screen shot but here are my results with my i7 930 at 2.8ghz

Time: 219.145 seconds
Speed: 35.1717 GFlops
Results: 3.188401e-002

I don't exactly know if it's good or bad. Based on my observations I'm in line with Directives CPU. His and PB's CPU's are both at least 1.2GHz faster then mine.

I guess if you scale it directly (and the performance probably wouldn't scale directly), my CPU at 4.0ghz would yield 50GFlops and at 4.5GHz my CPU would yield 56.25 GFlops which would still put PB at almost double my score.

I'm guessing that the AVX instruction set has a lot more to do with it then clock speed and efficiency per clock cycle.