Printable View
Well i got my venice prime stable at 2808mhz. One problem it needs alot of volts, 1.7v. I'm running this on a DFI, with a TT Big Tsnami cooling it. My max temp with small FPUs is 54C, idle at 36C. Am I gonna kill this cpu?
Also any idea to how many watts this thing is?
Thanks :D
Don't think you will kill it, any time soon atleast. The load temp seems in the high range. You might want to find better cooling.
Which ram and mobo are u using?
Perkam
CH-5, DFI Ultra DQuote:
Originally Posted by perkam
Well i dunno, but pumping 1.7v is alot for a 90nm chip, this is prolly a huge wattageQuote:
Originally Posted by koei
1.7 seem dangerous, I would only use that voltage with watercooling.Quote:
Originally Posted by AMD-me
Gah, yea when i turn it back down to 2700mhz 1.475v i never exceed 40C load. I think ill use my 2800mhz for benching :DQuote:
Originally Posted by hitmanx2
Ive been tryign to take down top ten x700pro heh
Not sooner than at least 5 years from now (and by that time you'll have another cpu :D) if your temps are within amd specified range. I had my winchester running at 1.7V for 6 months without problems. Just watch your temps. :D
1.7v is really high for a 90nm chip, best to turn it down for 24/7 use unless you plan to go the phase change route.
i say as long as the temps are low voltage doesnt matter.
but 54°C is pretty high, so either work on ur cooling or remove the IHS.
supplying high voltage can also kill a CPU on the long run
OOoo, removing the IHS and trying to mount a large HSF like the Typhoon on it are a recipe for disaster...
I only know of one person who has done that and thier temps didn't really get all that much better anyways, doesn't seem worth the risk IMO.
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.
Honestly, with good air cooling it could be pretty difficult to kill that CPU at voltage you mention. You just don't hear a lot about folks killing their CPUs, do you? Unless you do sth silly and help it die :D
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.
Gradual Death Syndrome. But then it only happens to .13u cpus like Pentium 4's. But then again even at 1.65v watercooled, .13u chips will die slowly
@mesyn191Quote:
Originally Posted by mesyn191
eeehhhhhh :lol: 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!
If managed to get 2.2v on your AXPM then you don't know how lucky you are...
I went through 3 AXPs and 2 AXPMs and not one of them was stable above 1.85v with good water cooling (Swiftech Quiet Power case modded with a Eheim pump and bigger 2x120mm fan rad) and I couldn't get either of them over 2.4Ghz stable...
2v killed 1 of the AXP 2500+ BTW too on air cooling, and many other people could hardly run theirs stable @ 1.8v due to heat....
I think your experience isn't common and shouldn't be viewed as the normal way to go about OC'ing.
well, if they can't do 1.8v stable due to cooling, they need better cooling. all of this is irrelevant if the temps are too high, as temps definately killQuote:
Originally Posted by mesyn191
Only 1 chip had poor cooling though, all the rest had watercooling. They didn't die but most of the time I could hardly even get them to boot, temps never got over 10C above ambient with that WC rig too BTW...
@mesyn191
I don't know about that being my isolated experience. I used to run a folding/docking farm made of 5 NF2/AXPM setups. ALL of them run 100% LOAD, ~2.55-2.6GHz 1.95v 24/7 day and night for about 6 months. I only have one of those AXPMs left, best I had, that I use now. It's rock-solid @ 2.1v/2.69Ghz with SP-97/92mm quiet fan. Mobo on the desk/temps LOAD up to 53C.
Anyway... We're getting waaaay to far away from topic...
To summarize... I would not hesistate to keep Venice @ 1.7v with good air-cooling for 24/7 operation :D ;)
bachus:
Like I said I think you're damn damn lucky then. FYI I've only _ever_ had one good overclocker, a AXP 1700+ that did 2.2Ghz on air. Even with phase change my OC's suck ass (currently have a A64 3500+ @ 2.74Ghz @ -12C (core temp) with my VapoLS)...
ive never had a chip die on me. ive pumped 2.0v's through axp 24/7 on air. and pumped like 2.1v through an old a64 for a while and no deaths.
Exactly. Pretty sure excessive voltage can damage the junctions in a transistor, regardless of temps.Quote:
Originally Posted by mesyn191
There would be wear and tear no matter what voltage. Nobody really knows how long it will take and it's not exactly against common sense that 1.7v is not as safe as 1.4v but then again, I doubt the chip will die on you due to wear before you get your next chip.
Well, I've a TBred 2400+ that has been running for more than 1 year at 2'1V 24/7 watercooled, still ocing like a mad.Quote:
If managed to get 2.2v on your AXPM then you don't know how lucky you are...
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.Quote:
Originally Posted by bachus_anonym
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.
so would you guys consider 1.6v safe for a venice?
yep
Actually I killed one xp-m when I gave it 2.3V. No problem with 2.1V however.Quote:
Originally Posted by bachus_anonym
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. ;)
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.
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!Quote:
Originally Posted by TurboDiv
Yup, it's dead.Quote:
Originally Posted by Jimbo Mahoney
Out of curiosity, how does one determine what "excessive" is?Quote:
Originally Posted by mesyn191
The problem here is excesive is a relative term.
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.
Well I turned over a new leaf with my ram. So I'm running 304x9 ram @DDR498 1.5-2-2-6 3.2v :clap:
And its more stable then the 2800 (alot cooler too :D)
Yes, prepare your post the the "Sudden venice death syndrome" thread now.
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?
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
anyone?Quote:
Originally Posted by Rezag1000
ull be fine :toast:Quote:
Originally Posted by Rezag1000
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.
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???
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 :stick: :eek:
How did you measure ? PM-me
yeah, I have been trying to figure out how to measure Vcore with a multimeter for ages!!
Don't PM, post in the thread so everybody else with a DFI nF4 know where to measure :)
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. :)
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.
well an axp can handle way more heat than 90nm A64 cores
according to amd the max die temp of bartons is 85°C while winnies should stay below 65°C
Quote:
Originally Posted by Jimbo Mahoney
that's exactly how it went when I killed my claw @ 1.8v :smoke: :bsod: :wasntme: :D
2600mhz 2.0v for this 2500+ mobile at the moment, 50c loaded on air... its got a while to go yet.