View Full Version : A higher OC takes longer?
I just stepped up my cpu from 1562 142x11 to 1610 140x11.5 and now its taking longer per frame on a native villian. Should I step it back down?
flounder43
01-18-02, 10:46 AM
There are other issues that may be slowing in down, like how much you are using the comp, and what other progs are running. Also, I always wait to get a good average of lots of frames before deciding what is faster.
I doubt an overclock would slow it down, btw.
Ill wait till tomorrow and see what happens between now and then. The frame time average is slowly going down now. Time will tell.
SickBoy
01-18-02, 12:35 PM
You slowed your FSB down by 2 MHz. I would say that's the culprit. But I wouldn't think 2 MHz on the FSB would make a lot of difference.
SB
lennytiger
01-18-02, 02:15 PM
if your multiplier is up and fsb up further if poss it'll sort your prob out
gsxr750
01-18-02, 03:22 PM
I don't think 2mzh make any different too since you OC it to faster speed.
I think it might be due to system configuration change or your processor do not have enough Voltage juice.
I was thinking the same thing so I just turned it back down to where it was. I will wait till I get a new psu since my 5v is already low.
lennytiger
01-19-02, 12:24 AM
Originally posted by Wyno
I was thinking the same thing so I just turned it back down to where it was. I will wait till I get a new psu since my 5v is already low.
How low is your 5V line??
check if you have a POT in your PSU, to up the voltage a bit.....
S_Wilson
01-19-02, 02:19 AM
Originally posted by flounder43
I doubt an overclock would slow it down, btw.
If his overclock has put him into a semi-unstable position, then it's very possible for the computer to fold slower.
Nothing new here, but use Prime 95 Torture Test for a few loops. If it doesn't turn up a Hardware Failure or Round Off Error once it's warmed up then you should be fairly stable. Run it about half an hour. My experience has been that I will turn up an error within the first few minutes if there is going to be an error.
P.S. Prime 95 is fairly sensitive and has been a excellent indicator of general stablity for my systems but is by no means 100%.
There is no pot in the psu to turn it up. The psu is a generic one that came with the case. As far as my 5v line is stays between 4.6 and 4.65 v according to mbm. I cant run Prime 95 at my current setting for even 30 seconds without a hardware error. Dont know what it is but the system is pretty stabe though.
S_Wilson
01-19-02, 12:22 PM
Originally posted by Wyno
There is no pot in the psu to turn it up. The psu is a generic one that came with the case. As far as my 5v line is stays between 4.6 and 4.65 v according to mbm. I cant run Prime 95 at my current setting for even 30 seconds without a hardware error. Dont know what it is but the system is pretty stabe though.
You might try to find out who the manufacturer is and see if you can get wiring diagrams. I don't know if that will be possible with a generic. In my enermax, I replaced a resistor with an adjustable pot for my 5V line, fortunately I found a thread which told me what resistor to replace.
In Prime 95, if you get and error then something is not quite stable. Could be to many cycles on the cpu, fsb or not enough vcore or vmem for your settings. Your memory timings may be to aggressive. Adjust your CPU down until you find a stable place, then work back from there, adjusting voltage and memory timings as necessary.
SickBoy
01-20-02, 09:02 AM
Originally posted by lennytiger
How low is your 5V line??
check if you have a POT in your PSU, to up the voltage a bit.....
This is NOT something good to do. On most PSU's, the 5V line fluctuates quite a bit between idle and load. If you do this, it may fix your 5v line but will more than likely cause other problems.
Generally, a low 5v line is indicative of a low quality PSU. The best thing you can do is go out and buy a good brand name PSU (Antec, Powmax,Enermax), of at least 300W.
I would highly suggest not screwing around inside a PSU, it has caps that can kill you.
SB
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.