How far can you push it - CPU & Components
  Tom's Guide Forums » CPU & Components » CPUs » How far can you push it
 




Word :   Username :  
 
Bottom
Author
 Thread : How far can you push it
 
More Information

After reading several forums containing about the cpu's electron migration, I was curious how far can one's cpu voltage be pushed and without hitting over the safe zone. Also I'm into overclocking a bit and wondering what's the voltage limits of my 840. I know that too much voltage will decrease the cpu's lifespan and some so severe that it will lasts only for weeks. I don't know how far can I increase my cpu's voltage and the maximum voltage I have attain is 1.4v and that only lasted for several minutes for I am paranoid it will cause my cpu to burst like a popcorn.

So how far can I push my cpu voltage and at voltage will it have a sever electron migration? I just want to know how far I can oc my cpu and that it could last me for a while as in a year or so.

Related Product

Register or log in to remove.

More Information

Wow, pretty nicely detailed answers. Thanks.

Before I was able to overclocked my 840 from 3.2 to 3.840 (20%) at 1.35v. Now with my new water cooling I can't do it anymore. I guess it must be the new pump that's causing voltage fluctuation in the vcore causing my pc not to boot or blank display. I did 1.4v again just several minutes ago and achieve a healthy overclock of 3.840. With that frequency it runs at 40c idle and 48-50c full load on BF2. Now I don't really need to run that game at the speed since it can easily perform at stock speeds. But if I do want to run my pc at that oc settings let's say for the upcoming game "Crysis" which would be more demanding. I run my games at a favor of both performance and quality, and so by next year I would have no choice but to overclock or upgrade. I guess it's cheaper to oc but being a soft-core overclockers myself it brings me concerns when overclocking the cpu at long periods of time.

And so if I overclock my cpu to 3.840 or 4Ghz at 1.4-1.45vcore with temps of 40c - 50c, would that be fine to do so for long term use? 8O

More Information

Thanks again Jack, now I'm going to oc it to 4Ghz @ 1.45vcore. Well, I'm going to have to upgrade anyways when Vista and DX10 arrives, so I'll just oc it till next year. 8O

More Information

From memory the point at which rapid/radical electron migration occurs for Smitherfield is 1.700V i.e. no cooling in the world would stop it
So your 1.4V is very safe.

What matters now is how much higher you can push with just 1.4V using the new cooling. Also it would be a good idea to measure the vcore directly from the mobo using a multimeter. VR-Zone might have it for your mobo.
2 things you've got to watch for, undervoltage and vdroop under load which are purely based on VRM(mobo) and not caused by the PSU.

I'll get back to your PM later....

More Information

Thanks. Well, I am running now at 1.4v oc. Still no luck at 1.35v. Temps are soaring up fast at 3.840Ghz its 38c-40c idle.

More Information

I trust you have mounted the block properly. Or if you just want to be sure, do the 'unlock, twist+press, grind, relock' method to ensure maximum contact.
As long as you can push higher, temperature is the least thing I'd worry about. (the rig will crash way before the temperature reaches too high)

And remember you'll get ~10C higher with Core Temp anyway.

More Information

Well, that's the thing. The Storms' water block mounting screws are not good. It's just a simple screw and nut components and could twist along when tightening the spring loaded fasteners. I placed some glue on it but it came loose again. Well, I'm gonna have to re-mount the block tommorow and see if I could get lower cooling.

More Information

Use a plier to hold the buttom screw from spinning.
Same happened with my SLK-900U's bracket.

More Information

Okay I did it. I tightened the nuts using a plier. I damn thing I almost total my motherboard. When I was tightening one of the nuts fasteners, I slipped the plier and scratch the board. It scratch the surface and expose one of them printed wire. I almost shit myself, and I thought I cut the printed wire and it's pretty hard to see if it's okay. I was twisting the nut since I'm too lazy to remove the motherboard which is a pain in the arse. Well I think it's fine.

As for the remounting I did it again and applied a fresh coat of AS5, after thouroughly cleaning it of course. I tightened the spring loaded nut fasteners all the way. I thought it would break my mobo. Weew! I always gets paranoid and have panic attacks every time I work with this rig. I need to see a therapist. Right now I'm getting 32-34 idle that is with 3.8Ghz @ 1.4vcore. Don't know the load temps yet. Hope it would be lower than before. Need to run Prime95 on it as well.

I also manage to place 3 more 120mm fans in my radiator. So now it's sporting six fans, 3 for pushing and 3 for pulling air. Hope that helps as well.

http://img174.imageshack.us/img174/8143/6fanradte4.th.jpg

More Information

Okay I got 37c-38c full load on BF2. Which is a good lower temp. :)

More Information

Yes, good.... NOW PUSH IT MORE!!!

More Information

I have tried 4Ghz which is basically the 840's limit on water cooling. Besides at 4Ghz it fails on Prime95 :( but it is able to run BF2 no problem.

More Information

Nice.

More Information

Which page?

Maybe we overclockers can learn something from those PDFs next time (this time C2D) instead of through experimenting. :lol:

BTW, using 1.55V would still allow a Smitherfield to live through the designed life expectancy.
1.700V gives nearly one yr of lifetime before the clock speed starts to degrade.

More Information

Page 2.5, Table 2-2.

Stating that my 840 has a max of 1.55v and minimum of -0.3v.

-0.3v? How could you get a negative voltage? :?

More Information