safe vcore & temps for Barton?

[sykopath79]

More 2500+ OC'ing questions...

What should be considered a safe maximum VCore & temperature? My motherboard undervolts VCore by 0.05V so I need to set it a little higher in BIOS to get the actual VCore I want.

Right now I'm running at 11x200, VCore set at 1.75 in BIOS which yields 1.68-1.71V actual. Temps top out at 50ºC under Prime95 torture test, 47ºC under a "normal usage" heavy load (e.g. a few hours of gaming, SETI@Home, etc.).

 

[bjc112]

Chances are you can hit 2.2 on stock Vcore..

If your are undervolting by .50, go ahead and start with 1.80 and stick with that. If you need more i suppose you can go with 1.85 and run a 1.8v.. I run my 1700+ @ 2.4ghz on 1.79v

Bump your Vdimm to 2.8v

Temps are safe up till around 90C..

If it's stable, it' safe.

 

[WobbleWobble]

Can't hit 2.2GHz with my XP2500+ on stock v-core.

I need to select 1.825V in the BIOS, 1.875 actual (it seems to bump everything up by .05V) to hit 3200+ levels.

 

[sykopath79]

I couldn't hit 2.2 GHz on stock VCore either, but I think this was due to my mobo undervolting. What is the stock VCore for the 2500+ anyway, 1.65V?

Well I'm running at 12x200 (2405.46 MHz in wcpuid) at 1.8 VCore (~1.75 actual), bumped my VDIMM up too since the board's undervolting has made my memory screw up a couple times. I probably have room to go even higher than 2.4 Ghz! w00t!

Although I haven't thrown Prime95 at it yet, so it remains to be seen how stable the current settings are...

 

[sykopath79]

OK, so it fell flat on its face once I tried Prime95. The first torture test errored out in a minute or two, the second one froze my system immediately upon running it. Bumping VCore up to 1.85 didn't help and I'm not sure if it's safe to risk going above that. So I've dropped down to 2300 MHz (11.5x200) @ 1.8V (~1.75 real) and it's stable so far now.

 

[bjc112]

Originally posted by: WobbleWobble
Can't hit 2.2GHz with my XP2500+ on stock v-core.

I need to select 1.825V in the BIOS, 1.875 actual (it seems to bump everything up by .05V) to hit 3200+ levels.

Yikes, back that down.

1.85 is pushin it.

 

[bjc112]

Originally posted by: sykopath79
OK, so it fell flat on its face once I tried Prime95. The first torture test errored out in a minute or two, the second one froze my system immediately upon running it. Bumping VCore up to 1.85 didn't help and I'm not sure if it's safe to risk going above that. So I've dropped down to 2300 MHz (11.5x200) @ 1.8V (~1.75 real) and it's stable so far now.

Keep it at 2.3 no need for that extra life threating voltage for 100mhz..

Nice OC.

BTW, Prime 95 always does that, if it errors the first test, its going to lock for sure the second time... Just better of to reboot and change the setting instead of getting a hard lock.

 

[sykopath79]

Can you get away with higher voltage if you have good cooling? Or is high voltage (>1.8V) in and of itself bad regardless of temps?

Reason why I ask: I've got it up to 11x216 now (2377 MHz) and it's running pretty well, but I bumped my VCore up to 1.85V (1.8 real). At the higher speeds (2.3 GHz+) it fails Prime95 but otherwise seems fine for real-world use.

 

[bjc112]

Originally posted by: sykopath79
Can you get away with higher voltage if you have good cooling? Or is high voltage (>1.8V) in and of itself bad regardless of temps?

Reason why I ask: I've got it up to 11x216 now (2377 MHz) and it's running pretty well, but I bumped my VCore up to 1.85V (1.8 real).

You got it, it's bad reguardless of temps. You start to break down resistors.

I would try and go as high as possible on 1.8 ( ACTUAL ) stop there.. No need to waste more life of that CPU.

 

[sykopath79]

Well, I couldn't get it stable above 2.2 GHz anyway... failed Prime95 in <1 hr, got a BSOD playing C&C Generals, hell Diablo II crashed and that doesn't even stress the system much. So I've gone back down to 11x200 @ 1.75V(actual) and it seems to feel at home running as a 3200+.

Was hoping to get higher than that, but oh well... certainly can't complain about saving $360 off a 3200+.

 

[VisableAssassin]

safe temp......keep it under 140f (60c)
safe voltage.......on air about 1.875 on water 2.0v
AMD states int heir tech docs that the barton can do .5v OVER stock so 1.65stock+.5v=2.15v however that is insane as hell

 

[bjc112]

Originally posted by: sykopath79
Well, I couldn't get it stable above 2.2 GHz anyway... failed Prime95 in <1 hr, got a BSOD playing C&C Generals, hell Diablo II crashed and that doesn't even stress the system much. So I've gone back down to 11x200 @ 1.75V(actual) and it seems to feel at home running as a 3200+.

Was hoping to get higher than that, but oh well... certainly can't complain about saving $360 off a 3200+.

Absolutley not.

 

[sykopath79]

Originally posted by: VisableAssassin
safe temp......keep it under 140f (60c) safe voltage.......on air about 1.875 on water 2.0v AMD states int heir tech docs that the barton can do .5v OVER stock so 1.65stock+.5v=2.15v however that is insane as hell

?! Are you sure about that? Sh*t, if I can get just 1.825V or 1.85V I should be able to stabilize it at 12x200. Needless to say I'm hesitant to jump on trying it though based on just one person's word when I've heard several remarks to keep it <= 1.8V.

I'm also wondering if perhaps my chipset voltage is needing a boost to make it stable at higher speeds. Seeing as everything else gets undervolted I wouldn't be surprised in the least. Which voltage setting controls the chipset? 5VSB? VBAT? I see those in Winbond Hardware Doctor and don't know what they reference. I think the +3.3, +5, and +/-12 are all reading the rails off the PSU but beyond that I don't know.