winchester, maximum safe voltage under air?

[jkresh]

Currently running my 3200+ at 2.5 with 1.56 volts, max load temp is 44c (XP-120 with panaflo ultra high speed), thinking about upping voltage a little more, but dont want to damage the chip, I know my temps are well within range but not sure how high a winchester can go before chip starts to take damage?

 

[cirthix]

as long as your temp stays low, you can keep adding juice

 

[beatle]

Bump for more opinions.

 

[cbehnken]

Seems the general thoughts are 1.6 is okay, but more than that are you are pushing it.

 

[jbh129]

extreme voltage is not for the faint of heart. I like to keep my cpu's right on the edge of tolerance. I believe that devices operate most efficiently just below the point of failure.

 

[Icepick]

Just finished reading this overclocking guide.
http://forums.anandtech.com/me...1072&enterthread=y
(This is the sticky at the top of this forum.)

In that guide it states that the maximum voltage the Athlons can handle is 1.8V. So even at 1.6V you're still OK as long as you can keep load temps under 55C.

 

[Acanthus]

I would not attempt anything over 1.65v

 

[frootbooter]

Originally posted by: icepik
Just finished reading this overclocking guide.
http://forums.anandtech.com/me...1072&enterthread=y
(This is the sticky at the top of this forum.)

In that guide it states that the maximum voltage the Athlons can handle is 1.8V. So even at 1.6V you're still OK as long as you can keep load temps under 55C.

That's for an athlon xp, not an athlon 64.

As long as your temps aren't too high I wouldn't be worried about if there's too much voltage.

 

[stelleg151]

I wouldnt go over 1.7.

 

[RussianSensation]

Originally posted by: cirthix
as long as your temp stays low, you can keep adding juice

No thats not a good answer. Even if you keep your processor as cold as possible voltage will eventually "kill" it as electron migration will wreak havoc. The general rule of thumb is not to exceed voltage for the cpu by more than 10% on air cooling.

This goes hand in hand with a P4 Northwood core for example. Stock rated at 1.48-1.55 => max recommended on low end is 1.63V (so round off to 1.65) and thus high end (requires water or non-air cooling means) of 1.70.

Now let's look at AXP: someone mentioned AXP has a safe voltage of roughly 1.8V. This once again goes hand in hand with the 10% figure since AXP is rated at 1.65 originally => 1.815 is safe.

Therefore, Winchester 90nm is rated at about 1.40 I believe (down 0.10 from 1.50 of the Newcastle core). So safe voltage should be around 1.54 (round off ~1.55 and most likely 1.60 watercooled).

Of course this 10% "rule" wouldn't apply to mobile products, but for desktop chips, it's a safe bet.

Also in a lot of cases, once the CPU reaches its stability ceiling after the first 10% increase in voltage, most of the time any additional increase in voltage has no effect on stability.

 

[Matrix333]

ive been running mine at 1.75v and idle temp is around 36C with load around 44C

am i okay?

 

[Matrix333]

ok, nevermind
load temp just hit 60C !!!!

while running prime95

im going to lower the vcord right now!

(these temps are accoring to MSI Core Center but id rather be safe than sorry)