• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

CPU scores below avrg. Broken after hard use?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
In reality it's not a huge difference they just improved on what was alredy there. They improved the die itself for the efficiency, better OCing and worked a couple bugs out. In day to day use most people won't notice that their Zip file opened a fraction of a second faster.
 
First up we don't really know much in this thread. So here are some useful things you could supply that are likely to get you more help.

Screenshot of HWMonitor (regular, not pro trial) with Prime95 running and having been running for at least 5 minutes and with whatever OC you plan on using applied. That'll give us an idea on temps and voltages.

Screenshot of CPU-Z's CPU, Memory and SPD tabs with your preferred OC applied. This helps us pick out any possible configuration issues.

More relevant/reliable benchmarks(passmark has a history of being weird). SuperPi 1M and 32M benchmarks / wPrime 32m and 1024m benchmarks / Cinebench R11.5 are some suggestions.
 
B-tec, I have a 955be which is basically and underclocked 965 if I'm not mistaken. If you run some futuremark benches that I'm more familiar with I can post you my results so you can compare the two. I can even set it up exactly as you have yours set, ie ram timings and overclock, so it will be a more accurate comparison.
 
Just for fun I downloaded passmark7 and ran a cpu bench and compared to another on the web that was at 3.9 heres the results.

Capturepass.PNG
My results were almost identical to the base I downloaded from the net. Also very close to the results you posted for the CPU @ 3.8
 
Last edited:
Just for fun I downloaded passmark7 and ran a cpu bench and compared to another on the web that was at 3.9 heres the results.

View attachment 124008
My results were almost identical to the base I downloaded from the net. Also very close to the results you posted for the CPU @ 3.8

Oh my gawd "Johan45", you beat that red thing everytime but that blue sucker whupped you every time. I bet you thought that cpu of yours was bad to the bone, but now we know it is beaten by that blue thing everytime. Guess you need to throw it in the garbage. I will send you address to garbage can. Hehehe. ROTFLMAO.
RGone...ser. :bang head:bump:
 
Nice one Rgone :thup::rofl:
But seriously I think the OP's issue is the C2 CPU. All his scores were good except the first few and that's where AMD made their improvements. The way a CPU handles it's memory would make that kin of difference. Just my opinion. :salute:
 
Mandrake, thanks. I'll probably check 3Dmark later. PCmark gives the results by opening a long, messy script with an internet explorer...
Johan, thank you. That was helpful. Doesent clarify the cause, but it definately tells I got a problem with the CPU. I think. The first 3 or 4, stats, of wich two first (maths) you got normal. I dont. For examle my integer math can be even 60% less!

Checked Cinebench 11.5 and here it is. Maybe someone can do that too with a same kind of setup?
123.png
 
The Cinebench CPU rendering score at 4.67 is what it should be. That's about what I get on my X4 955 overclocked to 4.1 ghz.
 
I felt so bad Rgone I had to do some checking and the blu guy had an advantage, I was only running @ 3.77. Forgot I was messing with things the other day so switched back and whooped his silly blu butt.


Capturepassnew@39.PNG


Ha no more garbage chip I think I'll keep it thank you!
Here B-Tec I did a run on cinebench and yours did better than mine, :bang head


Capturecinnew.PNG


Alot of factors will change a benchmark.
 
Flash your motherboard to an earlier BIOS. The first 1 or 2 bios released for the Sabertooth 990FX and the CHV were optimized for PH-II operation, later BIOS featured changes that improve performance on the BD/PD architecture but reduce performance / compatibility with older PH-II CPUs.
 
Last edited:
Fash your motherboard to an earlier BIOS. The first 1 or 2 bios released for the Sabertooth 990FX and the CHV were optimized for PH-II operation, later BIOS featured changes that improve performance on the BD/PD architecture but reduce performance / compatibility with older PH-II CPUs.

"Flash" it as well.
 
Whew what a relief. I thought I was going to have to pay shipping on the garbage can stuff. Good show man. :clap:
RGone...

I felt so bad Rgone I had to do some checking and the blu guy had an advantage, I was only running @ 3.77. Forgot I was messing with things the other day so switched back and whooped his silly blu butt.


View attachment 124021


Ha no more garbage chip I think I'll keep it thank you!
Here B-Tec I did a run on cinebench and yours did better than mine, :bang head


View attachment 124022


Alot of factors will change a benchmark.
 
Flash your motherboard to an earlier BIOS. The first 1 or 2 bios released for the Sabertooth 990FX and the CHV were optimized for PH-II operation, later BIOS featured changes that improve performance on the BD/PD architecture but reduce performance / compatibility with older PH-II CPUs.
Thanks for very useful info. Will look into that asap. Today maybe. I felt this chip was easier to OC on my previous MOBO: AM3 Crosshair III.
 
Flash your motherboard to an earlier BIOS. The first 1 or 2 bios released for the Sabertooth 990FX and the CHV were optimized for PH-II operation, later BIOS featured changes that improve performance on the BD/PD architecture but reduce performance / compatibility with older PH-II CPUs.
I think he meant the earlier versions like the second or third
 
Ok flashed 0.705 bios (second version available) after googling that it would be it. No difference in benchmarks. Conclusion: It must be the C2 vs C3 difference. Will anyway try this bios and see if its better.

And I forgot to thank Johan for the Cinebench result post :)
 
Hey NP B-Tec. We're always here to help :attn:
I've been doing the same this weekend, Testing some different BIOS versions for my 8350 to see if I can find a sweetspot, so far the results are iffy. I found a slight improvement in Vcore but nothing earth shattering yet, :bang head
 
Ok flashed 0.705 bios (second version available) after googling that it would be it. No difference in benchmarks. Conclusion: It must be the C2 vs C3 difference. Will anyway try this bios and see if its better.

And I forgot to thank Johan for the Cinebench result post :)

Thought I'd hook you up with a C2 comparison since not many people have them, although its a 955 basically the same cpu, I don't think its the C2 to C3 thats causing the problem, cpu at 3.9, timings on ram relaxed a bit though, 8-8-8-22 and cpu/nb at 2400. You got me just like Johan45 in cenebench too lol.

cinebench.PNG


pcmark.PNG
 
Tried Khan's setup as well as numerous setups in 3800-4000 range. All the same. Cinebench is only what makes sense. I feel like I must leave this behing, but keeo it in my mind. Maybe there will come an "aha" experience when I least expect it.

Crashed my so called stable setup that made thru few hours of everest stability test, but didnt make it one time in cinebench. Now have about the same hertzes, but fsb 200, instead of 230. Also found out "Probe Thermal" has no negative inpact on stability in my case. Cpu runs few Celcius cooler now. Also testrunning memory at 1600mhz, wich Phenom II C2 Seems to handle after all. 1540mhz with CL7 was 5% faster, but that setup wasnt in fact stable.

Did you peeps know Sabretooth 990fx (r.1) cant keep its FSB tight at what it should be? My FSB wanders +1% and back. So Fsb 200 is actually jumping from 200 to 202 or so. I dont think its never 200.0 or under that. This is normal?

With 200fsb now:
CPU 3.9Ghz
NB 2800mhz
Ram 1600mhz
 
Back