Am I gonna kill my venice?

[J-Mag]

How do you guys know if 1.7v is deadly for 90nm Winchester/Venice? I've never run my Winchester @ below 1.7v, even on air. That 54C in SmallFFT is OK, so I would keep it If you can get better temps by changing cooler, than do it. If not, keep it as it is.

Agreed: Ever since I got my DFI I have been running my winchester at 1.7v on an XP-90C. Honestly I could care less if the thing died on me, but if i was running a $1,000 proc I might be a tad more worried.With my winchester I see a max of 48C or so under small FFT, while my venice pushed ~54C under 1.7v with the same cooler. Interestingly though my Venice only held out for 5 days vefore I got the 4 LED prob ( I had been priming my venice 24/7 at 1.7v ~2.67ghz ). However I also noticed that one of the XP-120 fan mounting clips somehow got on the board right before the 4 LED happend, so most likely it wasn't the voltage that killed it, just a short.

[Jimbo Mahoney]

Just to let you guys know:

I'm 99% certain I fried my venice with 1.72 volts (1.7 actual).

I'd been running it for ages @ 1.65 actual and it was getting better and better. I decided to push it again and went up to 1.72 volts in the BIOS.

Save and exit and computer just turns off.

OK, turn it on again. It starts for a split second, then powers down.

Anyway, long story short, new CPU arrived today and I'm up and running again.

I'll test the old chip again later to be sure.

Real bummer 'cos this new chip sucks so far. I managed 2.9 Ghz Prime stable on the old one (now RIP). This new one is struggling to hit 2.6 Ghz

Oh temps were always good 'cos I have an excellent watercooling rig.

Highest I saw them go was 48'C, but that was with the chip putting out 160 watts.

[Rezag1000]

so would you guys consider 1.6v safe for a venice?

[ReelMonza]

yep

[kiwi]

@mesyn191

eeehhhhhh I stopped reading those 1.5yrs ago, when I gave 2.2v to my AXP-M 2500 and had it 24/7 @ 2.1vcore. Everybody around was saying that > 1.85v on air = DEAD!!! I still have that CPU, in my everyday/work NF2 @ 2690Mhz / 2.1vcore

Of course, it's all truth. It will die one day, but question is WHEN!

Actually I killed one xp-m when I gave it 2.3V. No problem with 2.1V however.

[TurboDiv]

All you guys need is max 1.5 years of life for a CPU. After that period it is pretty much outdated and worthless for top/mid gamers... My lame Winchester is at 1.66 Volts on Xp-90 with TT silent Cat fan and i've seen it as high as 57 degrees in the hottest days but it is working. Bought it like 9 months ago.

[higgins]

well ive run my XP2400+ at 2,1V air cooled for around 1 year... it's starting to crap out now. but i mean, that's to be expected.

[DilTech]

All you guys need is max 1.5 years of life for a CPU. After that period it is pretty much outdated and worthless for top/mid gamers... My lame Winchester is at 1.66 Volts on Xp-90 with TT silent Cat fan and i've seen it as high as 57 degrees in the hottest days but it is working. Bought it like 9 months ago.

The A64 came out more than 1.5 years ago.. Is there any A64 that is "worthless for top/mid gamers" when overclocked?... In the past year and a half all we've done is gone up 200 mhz and added dual channel.... The Clawhammer 754's still kick rather hard in any desktop!

[Jimbo Mahoney]

I'll test the old chip again later to be sure.

Yup, it's dead.

[J-Mag]

http://www.realworldtech.com/page.cf...WT122199000000

"Excessive supply voltages overstress the transistors and can lead to damage due to charge injection from the hot electron effect or damage to the ultra thin gate oxide dielectric from excessive electric field strength. Even if the stress is insufficient to lead to an immediate catastrophic failure, the accumulation of low levels of damage can lead to a drastically shortened device lifetime."

Its a ooold article but all the info. is still dead on.

Out of curiosity, how does one determine what "excessive" is?

The problem here is excesive is a relative term.

[Jimbo Mahoney]

I personally find that components have a tolerance of +20%.

I would say more than 20% over spec is execessive.

1.4 x 120% = 1.68v

Therefore, I was definately using excessive voltage.

[AMD-me]

Well I turned over a new leaf with my ram. So I'm running 304x9 ram @DDR498 1.5-2-2-6 3.2v

And its more stable then the 2800 (alot cooler too )

[ls7corvete]

Yes, prepare your post the the "Sudden venice death syndrome" thread now.

[Rezag1000]

ITE smart guardian and CPU-Z says my CPU is at 1.58V, and I have it set to 126%x1.3 in BIOS.

Should I expect my CPU to last more then 2 years?

[Peen]

Man I hated my XP1700 Tbred so much, I put 2.4v through it on air. Didnt kill it though, ran 2.2v 24/7 to get 2.3ghz. I had bios options up to like 2.65v I think.

If you can replace your chip no probs, then I would run 1.7v 24/7. If it was a FX55 or something I would probly turn it down =P I cant remember the last time I heard of a chip dying from too much voltage over time, usually just OC a little worse

[Rezag1000]

ITE smart guardian and CPU-Z says my CPU is at 1.58V, and I have it set to 126%x1.3 in BIOS.

Should I expect my CPU to last more then 2 years?

anyone?

[AMD-me]

anyone?

ull be fine

[fareastgq]

I'm planning to run my sandy at 1.75, heh. alrdy tested it, it won't die. Maybe I'll even pump 1.8 into it, grin, if it dies, oh well, I'll just get an x2. I think your fine at 1.7 as long as your temps stay down. The best trick is better cooling.

[Zebo]

Everybodies guessing. AMD never told us "safe" values for A64's in their white papers like they did with socket A chips so technically you're risking anything above 1.4 volts.

I am persoally comfortable 10% over stock Vcore.. dunno why..just a nice number...10. Who knows maybe able to hang with 2.2Vcore like bartons could???

[Jimbo Mahoney]

Just be careful guys - I just mesasured my Vcore using a multimeter on my DFI Ultra D.

Software says 1.648

Multimeter says 'uh uh' - 1.73

[ReelMonza]

How did you measure ? PM-me

[Rezag1000]

yeah, I have been trying to figure out how to measure Vcore with a multimeter for ages!!

[exscape]

Don't PM, post in the thread so everybody else with a DFI nF4 know where to measure

[kemist]

You guys should also keep in mind that as the process gets smaller more of the wattage is lost through leakage, so while 1.7v might sound like a lot for 90 nm a lot of the extra power may just be leaking through the processor and not doing any actual work, im not certain if this is true or not but leakage definitely increases. So basically what im trying to say is while increased voltage will def kill the processor eventually, it may be that with a smaller process as long as the die is kept cool, the increase in voltage may not be as significant as it was on older larger process processors. If i am just talking out of my ass let me know.

[uscfan]

You'll be fine. Just keep below 60 and you really dont have anything to worry about. These things can do above 70-80C really.

Hehe a friend of mine had his A XP reboot with this loud popping sound once, and he found out it was running over 100C for a few minutes. Really weird, but his survived and still hits 2.7 just fine.