It looks like you're new here. If you want to get involved, click one of these buttons!
If I overclock my 3.3GHZ Black Edition 1100T, To 3.8GHZ (Seems fine) But is it fine longterm? ASUS Tune V EVO Stated that its the max stable overclock for my system.
Few questions.
Is there any long term damage to that? I noticed even at 3.3GHZ my voltage goes from 1.4 to 1.475. However with the overclock, the overlock voltage is constant 1.4. Now, shall I go for the overclock and keep it that way, (with cool&quiet off) or shall I leave it stock? I tried to push 4GHZ but its not stable.
Do you think if i pumped more juice into it at 4GHZ I could get it to be stable? It was using 1.4 I would try the 1.475.
Thoughts?
Comments
Note that even at stock speeds, your processor will clock up to three cores up to 3.7 GHz if under heavy loads and the other cores are idle. You'd need a substantially better motherboard and processor cooler for it, but running the cores at 3.7 GHz all the time is only nominally overclocking.
In general, you'd prefer to run your processor at the lowest voltage that you can. If you're overclocking the system to a fixed maximum clock speed, and not making the speed fluctuate with turbo core, then you can mess with voltage to see how low of a voltage is stable. If you set the voltage too low, the system will be unstable because the processor will sometimes return the "wrong" answer to computations, but that won't damage the processor. Setting the voltage too high can fry it, though.
Power consumption is proportional to voltage squared, so if you can reduce the voltage by 10% (which is actually a lot), that reduces power consumption by 19%. That likewise reduces heat output. Lower temperatures can also reduce power consumption in some cases. At stock speeds, this isn't a big deal, but if you're overclocking, it can be nice to bring power consumption as low as you can at a given clock speed, as that means less stress on all hardware involved.
There's also a question of how stable a processor is. A processor that seems to work fine most of the time, but crashes under heavy loads, isn't really stable. You can test with Prime95, with one thread running on each core, as that will push your processor very hard. If a given overclock is stable under Prime95, then it should be stable in just about everything else.
Higher voltages do typically let you get higher clock speeds, so long as temperatures don't get out of hand. But if you're planning on leaving the processor at 4 GHz and 1.475 V indefinitely, even if it's stable for a while, that's running a pretty big risk of frying it.
i should give you more information really
the psu is ocz 550w fatal1ty
the board is m4a89gtd pro http://www.techspot.com/review/252-asus-m4a89gtd-amd-890gx/
the cooling is Arctic Cooling Freezer 7 Pro rev 2 Socket 775, 1156, 1155, 1366, AM2, AM3 Heatpipe CPU Cooler
Seriously, are you people high? Overclocking even a dual core that's at 3ghz is just stupid and a waste. There is NOTHING out there right now or probably in the next year that even NEEDS 6 cores let alone even a fully powered quad core.
When did you start playing "old school" MMO's. World Of Warcraft?
There's also a question of how stable a processor is. A processor that seems to work fine most of the time, but crashes under heavy loads, isn't really stable.
This ^
Also im not sure what to do with my memory, the auto overclock runs the memory at like 950 or something with the cpu being at 3.8, my memory is 1333mhz stated.
i do usually use turbo core to activate the 3.,7ghz on 3 cores as stated. this is normal.
i can boot into a 4ghz, but i have my multiplier set too high and the memory at 1500mhz which is too high also so it is not stable when i run the PI test calculation it crashes after 60 seconds. cpu at 1.4
Im not sure what to change to make it stable. kinda tried all kinds of options, also a multiplier of 20 doesnt work so well i noticed, it likes to be around its default of 16.5 the Be has an unlocked multiplier, so there are too many variables to consider for me to work out what the issue is exactly, im not sure, if i need to give the northbridge more power,.. if i change the NB at all in the bios even a 0.20 increase the text for that system turns red. so i dont bother to try,
there is a guy who did have the 1100t stable with the same mb however with 1600mhz memory and he bumped the nb to 1.4
im not sure whether to try that or not
it's called overclocking and enthusiast... its kinda what we do, -- meaning we push our system to get the max out of the components.. thats why i purchased a black edition, its made for overclocking, etc.
4ghz is the golden number these days, everyone wants to hit it. i am just trying is all. need advice, so thats why i am posting.
The motherboard should be fine with however you want to clock the processor, within reason. It's not what you'd want for liquid nitrogen, but it shouldn't hold you back on air cooling.
The power supply is dicier. OCZ is not a good power supply brand, and their Fatal1ty lineup isn't just similar power supplies with different wattages. The 750 W version is made by Highpower, the 700 W version by Impervio, and the 550 W version by Sirtec. Hardware Secrets found that the first two of those are reasonably good, but that doesn't say anything about the 550 W version, which is what you have. It often isn't a fluke that companies send their better products in for reviews and not their worse ones.
Arctic Cooling says that your processor heatsink should be able to handle up to 130 W just fine. Your processor has a TDP of 125 W at stock speeds. Now, Arctic Cooling is really just being conservative with their 130 W rating. But it's not the sort of heatsink you'd want if you're looking to push the processor as far as it can go. 3.8 GHz at the stock voltage or lower should be pretty safe so long as the CPU temperatures that you measure directly are safe. But I wouldn't want to overvolt the processor to try to be able to clock it higher on that cooler.
The Phenom II X6 1100T is a black edition processor, which means it has an unlocked multiplier. You should be using the multiplier to overclock, and not adjusting the BCLK unless you're trying to squeeze out every last drop of performance. Which you shouldn't be. If you're only changing the processor multiplier, then that shouldn't change the memory clock speed.
the multiplier at 17, (up from 16.5 ) does change the memory mhz to 1400. up from 1333
set stuff back to stock just now, rebooted with a 20x multiplier, the mem is either at 1333 or 1400 i cant remember, but nothing in that regard was changed. the processor is unrestricted by cool & quiet now and is running at 4013.36 and im going into a game of BC2 to test is it crashes. brb
Do note that games probably won't push your processor that hard, as they're usually constrained by the video card rather than the processor, and don't scale well to six cores anyway. That's what a stress test like Prime95 is for. It can give each core its own dedicated thread, and also manages to push each core very hard by keeping it very busy.
Don't be afraid to increase the voltage a little, just do it carefully. Increase one step and run some stresstests, if the CPU doesn't get to warm, you can increase it a little more. If you run into stability issues when overclocking, increasing the voltage probably solves it.
The processor is also likely to suffer a little extra wear and tear from overclocking, but unless you go to extremes you will probably not notice it until it's time to upgrade anyway.
You should also note that overheating isn't the only thing that can fry a processor. A fancy cooling system that keeps the processor below room temperature at all times won't prevent it from dying of electromigration if the voltage and clock speed are too high.
How far to overclock is largely a question of how much it would bother you if the system isn't entirely stable, or if the processor dies entirely.
True, but usually the system will refuse to boot or you will run into some other problems (like extreme stability issues) before that happens, which is the reason to do incremental increases with stresstests inbetween. The cpu is very inlikely to fry going from 4000mhz and 1.4v to 4100 and 1.475v :P
Everything up until now has been unstable. Final results are:
3840.60MHZ / (232.76 X 16.5) Across 6 cores.
Mem is at 925MHZ, 8Gb DDR3 2GB+2GB+2GB+2GB =8 Unganged
As for Timings - not sure.
Volts to CPU are 1.4000v - Constant
Ran a PI Calculation Test on CPU , Then tested MEM, @ 150 Seconds Per component at 100% load - Passed Both tests. Stable.
Cant get any more out of my system I think. Everything up until now has been unstable. Final results are:
3840.60MHZ / (232.76 X 16.5) Across 6 cores.
Mem is at 925MHZ, 8Gb DDR3 2GB+2GB+2GB+2GB =8 Unganged
As for Timings - not sure.
Volts to CPU are 1.4000v - Constant
Ran a PI Test on CPU and MEM, 150 Seconds Per component at 100% load - Passed Both tests. Stable.
Cant get any more out of my system I think. Everything up until now has been unstable. Final results are:
3840.60MHZ / (232.76 X 16.5) Across 6 cores.
Mem is at 925MHZ, 8Gb DDR3 2GB+2GB+2GB+2GB =8 Unganged
As for Timings - not sure.
Volts to CPU are 1.4000v - Constant
Ran a PI Test on CPU and MEM, 150 Seconds Per component at 100% load - Passed Both tests. Stable.
Cant get any more out of my system I think.
As you know, different chips behave in their own way, mine, seemingly does not like 20x multipliers, )i did that manually so actually i probebly didnt give the other chips more juice to compensate, it does boot into windows and is stable at 4ghz on 20x multi, but i cant really do anything with it. granted none of the other options where changed, they were on auto. and that's possibly the problem.
was thinking about ocing my gtx 570 sonic platinum, but its actually factory overclocked so i cant see much point in that tbh.
I get 60 frames at max setting in 95% of games really, so im not sure why i am filddling so much. i just like seeing what my hardware is capable of i suppose ^^
To respond to the question of if overclocking will reduce the lifespan of the part:
The short answer is: Yes, but not appreciably so.
Bit longer answer:
CPU's are totally electronic parts: they have no mechanical parts, it's all just a lot of transistors. So they are pretty darn reliable. But they aren't foolproof. The hotter that an electronic device runs, the more likely it is to have some sort of electrical failure (trace blows open, transistor junction leaks, etc).
In overclocking, you are going to produce more heat than if you don't. Doesn't matter what cooling solution you use: an overclocked processor will run hotter than a stock clocked processor, if they are both using the same heatsink. And even if you do upgrade the heatsink, there is still the matter of thermal transfer: the heat actually has to get from the transistors, to the die, through the heat spreader, and then to the heat sink, and you can only see the temperature at specific points, so you can't really just go off what the thermocouple on the die says, because it can't read the temperature of the entire chip, just of the one specific area in which it physically is located.
Now, the question is, how much does the lifespan of a electronic component get reduced? The answer is that it varies and is somewhat random. In general, a CPU can last dozens of years, and there are no moving parts. Even at stock speeds there is a chance that it could die at any moment (however small). Overclocking just increases that chance somewhat. So even overclocked, it's likely to last years and years.