• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Undervolting performance testing

Wing-Man said:
thats about 25 years ago.
Click to expand...
yes, hence stating up till that point thats what we used and it worked for stability testing... That's why I'm genuinely shocked it's different now?

Wing-Man said:
there is no one test that dose it all. you need to test based on your use. if all you do is watch your tube and play games why do you use a test that hits the cpu with a constant heavy load, when want you need is constantly changing loads.

I use 3D Mark, RealBench, Cinebench R20 and aida64. i run them by there self in a loop and or more than one at the same time.
For ram you cant beat MemTest86 end off but ram dose one thing and only one thing. unlike a CPU that is constantly
changing work loads and types.

the first stress test is dose it post. but just because it posts doesn't mean it games.
when we are overclocking i keep bumping the multi until the system wont load to windows, that in itself is a test.

most CPU's will throttle when running prime if your not maxing the core how do you know its stable? then you open chrome and the cores hit full boost for 1 second and the system BSOF.
Click to expand...

All I know is up till my last system a q6600... all everyone used was prime95, 3dmark and actual usage to test it... I'd never ever heard of a overclock being unstable from opening a browser that was stable in 3d mark/prime95/gaming...
crazy. not at all saying you're lying just am genuinely shocked as that is literally what everyone used to use, as you say real world testing and then prime95 and 3dmark for ott security.
 
keef247 said:
yes, hence stating up till that point thats what we used and it worked for stability testing... That's why I'm genuinely shocked it's different now?



All I know is up till my last system a q6600... all everyone used was prime95, 3dmark and actual usage to test it... I'd never ever heard of a overclock being unstable from opening a browser that was stable in 3d mark/prime95/gaming...
crazy. not at all saying you're lying just am genuinely shocked as that is literally what everyone used to use, as you say real world testing and then prime95 and 3dmark for ott security.
Click to expand...
The way CPUs behave these days is much more different than the past.

For the Q6600 at least, it doesn't have turbo boost, only speed step, and even then speed step is pretty basic compared to the c-states we have now.

Nowadays we have CPUs that can switch off cores completely, and boost very high close to the limits. And they switch between these states much faster than the Q6600 did. This is why we need different sets of tests since Prime95 alone won't be able to stress all these points, especially the moment the CPU suddenly changes to a different state with a sudden rise/drop in voltage.

Personally I use a mix of Cinebench, AIDA64 and 3DMark first to see if there are any issues, then I proceed with prime95 to check peak temps, and finally CoreCycler to properly check that each core is completely fine. May be over the top but it gives me a better peace of mind that the CPU is boosting/downclocking properly.
 
eddiew said:
Well, 15, but it's still a long time in this sector :D

I remember envying a housemate when he got a Q6600... and now I feel old.
Click to expand...

i never had one, went C2D to AMD phenom ii

keef247 said:
as I thought lol
Click to expand...

the one reply the supports you... ** Please try not to bait other users - EVH **. the program is trash, occt uses the same testing method but in a more useable way but it still only hits one type of load and needs many other programs to show stable.
people say use it to test for heat.. WTF WHY?... ho no under a stupid load my CPU hits 91c but when i just use it normal 55c so in what world do you need to know about the 91c.
So many people start prime and go to bed, waking upto a dead PC. yes that a program i can get on board with.
 
Last edited by a moderator:
Re undervolting

The ultimate test is just to use the PC. Everything undervolts, chipset, CPU, memory. You have to adjust one thing at a time in small changes, then see if the computer is stable, however it can sometimes take a week before you know if a new voltage is correct.

Also the better quality PSU the more you can undervolt.
 
JasonM said:
Re undervolting

The ultimate test is just to use the PC. Everything undervolts, chipset, CPU, memory. You have to adjust one thing at a time in small changes, then see if the computer is stable, however it can sometimes take a week before you know if a new voltage is correct.

Also the better quality PSU the more you can undervolt.
Click to expand...
fair enough mate, I shall do that and run all of the aforementioned apps above.
psu wise I'm thinking a Corsair rm850w gold fully modular? any good? seems reasonably priced?

Orcvader said:
The way CPUs behave these days is much more different than the past.

For the Q6600 at least, it doesn't have turbo boost, only speed step, and even then speed step is pretty basic compared to the c-states we have now.

Nowadays we have CPUs that can switch off cores completely, and boost very high close to the limits. And they switch between these states much faster than the Q6600 did. This is why we need different sets of tests since Prime95 alone won't be able to stress all these points, especially the moment the CPU suddenly changes to a different state with a sudden rise/drop in voltage.

Personally I use a mix of Cinebench, AIDA64 and 3DMark first to see if there are any issues, then I proceed with prime95 to check peak temps, and finally CoreCycler to properly check that each core is completely fine. May be over the top but it gives me a better peace of mind that the CPU is boosting/downclocking properly.
Click to expand...

Yeah I remember we used to have to turn off speedstep altogether haha. Yeah I'd always run prime95 first the whatever 3dmark was out at the time and then just use the computer and see what happened and adjust accordingly.
Ah nice list of things to try then, nice one for that. I've been too busy playing with drift cars for years and neglected the overclocking etc scene for awhile but I'm looking forward to getting back into messing around with pc's and tweaking stuff again. It's interesting to me this time around as I'm not doing it for performance but for lower tdp's and potential prolonged performance without throttling, the temps side doesn't really bother me in turns of it not kicking out as much heat but it is a bonus!
 
Last edited:
Wing-Man said:
occt uses the same testing method but in a more useable way but it still only hits one type of load and needs many other programs to show stable.
Click to expand...

This is false.
unknown.png


That's a full array of tests within occt you can use. I assume you just latched onto my comment about occt making P95 tests easier to consume. You clearly don't have much experience in stress testing software, use cases and outcomes. So why continue to offer advice? I don't stand infront of a GP office and offer medical advice for the same reason.
 
Chaps, just a reminder... there's no need to be so passive aggressive toward each other. Let's discuss this in a civil way please, thanks.
 
You must log in or register to reply here.
Back
Top Bottom