Overclock test
- Thread starter moremoney
- Start date
At any of the below options, really:
- That you actually asked, which must mean...
- You didn't read the OVerclocking sticky at the top of this forum or...
- Use the Search function to have a look at any of the other hundreds of threads on subjects pretty similar to this one or...
- You're only using Intel Burn Test to determine whether or not your system is stable (it might be stable through IBT, but suffer epic failure in a game...) or...
- The fact that you've only overclocked the bus by 66Mhz and are actually testing an E8500...
I'm sure more useful contributions will come along in time...
At any of the below options, really:
- That you actually asked, which must mean...
- You didn't read the OVerclocking sticky at the top of this forum or...
- Use the Search function to have a look at any of the other hundreds of threads on subjects pretty similar to this one or...
- You're only using Intel Burn Test to determine whether or not your system is stable (it might be stable through IBT, but suffer epic failure in a game...) or...
- The fact that you've only overclocked the bus by 66Mhz and are actually testing an E8500...
I'm sure more useful contributions will come along in time...
Are you for real?.
And i am being polite.
Numbers help dots mean ****e.
1. yes i read the sticky.
2. I have been hear long enough, and have not seen my question asked the same way as mine. And who cares......... if you don't want to give an answer DON'T.
There are plenty of nice people on here who are only to willing to give a positive response.
3. Duh. Thicko, that is the reason for my question.
4.FFF, 66mhz? Put your goggles on and read again.
The post was meant in jest, as it's late, but...
I think you'll actually find that most of my post contained very useful advice... reviewing existing posts, and using the 'search' function.
IBT is a very specifc test: it will tell you if your CPU is calculating correctly. It will not tell you if the overclock will work well when the whole computer is given a workout and different levels of stress over a period of time.
Any overclocker that just uses Prime95, just uses Orthos, just uses IBT is very naive when testing for 'stability' as each tool only explores one facet of a computer's performance. On the flip side, if the computer works as you want for what you want, then what is the problem? Why bother with running through all these tests?
And yes, your BUS is only overclocked by 66MHz (this is explained in the sticky...). Just to explain where that number came from...
Stock Intel E8500 -> 333 (front side BUS) x 9.5 (multiplier) = 3163 (ish) MHz (CPU clock speed).
Therefore, an increase of 333 -> 400 (as your original post indicates) would require an additional 66MHz on the BUS.
Last edited:
Thank you, that is a much better answer than your first one.
Considering the first half is essentially a copy'n'paste from any other number of threads where I've tried to explain the concept of 'testing for stability' in response to posts that are very similar to yours, I would think you went through a lot of effort and patronising to get a half answer.
Now go back, read the sticky properly, search for "OVERCLOCK WOLFDALE" and kick your CPU's ****. E8500s should be good for around 4.2GHz without too much trouble...
I think you missunderstood me. i have clocked my chip to 4.0 on my Gigabyte board before, but my dog ate all my notes. Honest.
All i ask is, is this one test good enough for stability because i am an impatient sort of chap and hate running prime all night and day.
Again.
I thank you for being so patient with one so ignorant.
Last edited:
Don't thank me (even if you're being sarcastic) - do some reading.
Quick question: does your computer crash when doing what you want it to? No? Then it's probably stable enough.
As the sticky says, testing will not prove stability, but will exclude instability for a certain set of functions (i.e. running your favourite game/watching HD videos/encoding stuff/etc.) assuming you test them directly (play a game for 10 hours without issue and chances are you computer will be stable playing that game for any given period of time).
The risks associated with instability can be pretty huge and entail nasty consequences, but they are only risks: not guaranteed outcomes.
If running prime while your asleep is annoying
But having software crash is not
Then leave it be and call it good.
I consider 5 runs of ibt shamefully inadequate to test an overclock. It's taken less time than this thread has, surely thats a sign that you're not being thorough?
But having software crash is not
Then leave it be and call it good.
I consider 5 runs of ibt shamefully inadequate to test an overclock. It's taken less time than this thread has, surely thats a sign that you're not being thorough?
I am old and i am tired. I have read and read page after page on 1/2 dozen forums when i first started to overclock.
Maybe i have got lazy, i don't know, i just can't do it all for every new spec board or processor, with all of these new settings.
Thanks for your help anyway. (no sarcasm intended).
Maybe i have got lazy, i don't know, i just can't do it all for every new spec board or processor, with all of these new settings.
Thanks for your help anyway. (no sarcasm intended).
But i have had it crash within the first hour, then my computer restarts and is left idle for 5 hours wasting energy.
I am\was looking for a stability test that was quicker than prime.
Spammer!
ibt is quicker than prime. 5 runs is hardly a lot though, and took under 5 minutes to finish. I tend to do 20 runs while testing as I clock, then prime overnight when its passing 20 runs of ibt. Electricity cost for 5 hours of idle doesn't bother me much.
If you're only going to use one test, that's probably the best one to choose (more than 5 loops of it). However I wouldn't consider passing a single test to be stable. I would certainly not consider passing 210 seconds of a single test stable, did it even reach equilibrium temperature in that time?
ibt is quicker than prime. 5 runs is hardly a lot though, and took under 5 minutes to finish. I tend to do 20 runs while testing as I clock, then prime overnight when its passing 20 runs of ibt. Electricity cost for 5 hours of idle doesn't bother me much.
If you're only going to use one test, that's probably the best one to choose (more than 5 loops of it). However I wouldn't consider passing a single test to be stable. I would certainly not consider passing 210 seconds of a single test stable, did it even reach equilibrium temperature in that time?
Spammer!
ibt is quicker than prime. 5 runs is hardly a lot though, and took under 5 minutes to finish. I tend to do 20 runs while testing as I clock, then prime overnight when its passing 20 runs of ibt. Electricity cost for 5 hours of idle doesn't bother me much.
If you're only going to use one test, that's probably the best one to choose (more than 5 loops of it). However I wouldn't consider passing a single test to be stable. I would certainly not consider passing 210 seconds of a single test stable, did it even reach equilibrium temperature in that time?
Did what reach equilibrium temps?.
Do you mean cores?.
Processor tends to get hotter under load, quickly then more slowly until it levels off. I meant the processor as a whole, though the cores should hit an equilibrium too.
Basically 3 1/2 minutes is way too short a test imo
Then i think i have a good chip.
One is stuck at 38 but as soon as the other hits 38 they move together. More or less.
I have read the e850 thread and i thought i knew now't.
I don't think i will be asking again.
Never ask people who pretend they know.
Now that is rude.
Rate of heat dissipation is proportional to temperature gradient. So the rate at which heat moves from heatsink to atmosphere is low when just starting, as the heatsink is cold. Therefore the chip gets hotter fast. Once the heatsink starts to warm up, it throws out heat faster. After a while the chip and heatsink hit equilibrium, at which point you hope the chip is under 70. It's often called Newtons law of cooling, and it's a poor approximation but it is easily accurate enough here.
Beyond that, form your own notion of stabilty and work to that.
Rate of heat dissipation is proportional to temperature gradient. So the rate at which heat moves from heatsink to atmosphere is low when just starting, as the heatsink is cold. Therefore the chip gets hotter fast. Once the heatsink starts to warm up, it throws out heat faster. After a while the chip and heatsink hit equilibrium, at which point you hope the chip is under 70. It's often called Newtons law of cooling, and it's a poor approximation but it is easily accurate enough here.
Beyond that, form your own notion of stabilty and work to that.
Last edited: