Printable View
A certain someone will not let up on thinking voltage kills them and insists I will fry all my chips. I've run all my A64s (a claw, venice, 2 diegos, and a diego opteron) to to 2v for extended periods of time and they all run just as good as always. So has anyone killed a chip from voltage?
ive seen my current chip degrade a little after running 1.9 volts for a few months...
DFI and Oskar Wu have.
Sorry, couldn't resist :)
I think its pretty well common opinion though that running that sort of voltage is definately going to shorten the lifespan, this is xtremesystems.org though, you break it and you buy something newer and hopefully better :D Getting a new goodie is half the fun of it.
i killed about 15..
and an additional 3 in the process of trying to unlock them...
Wow, so what'd you do?Quote:
Originally Posted by nn_step
just go to the A64 unlocking thread for the 3
The others were simply too high for too long...
Hence my current chip and it's overclock is at 1.4v
I could go farther but I love this chip...
You must really have some "special skills" to be able to pull off something like that :lol:Quote:
Originally Posted by nn_step
I've had maybe 15 different AMD CPUs (4 A64) and have never killed even one from overvoltage :) All were seriously "vcore-tortured", on a daily basis.
I only killed one A64 (Winchester) but that was because I was stupid :D It died at +95C @ 1.7v :lol:
What voltage were these chips run at? What cooling was used, what were the max load temps, and how long before they died? If you don't mind sharing of course. If you have lost that many chips, you should share your experience so we can avoid the situation.
NO... then again I haven't had the same one for more than 2 months. :p:
I did kill one with homemade phase kits, too much mounting hardware created a short I guess. Speaker wire from the mobo melted and other notices. The mobo and the newcastle died together. They smile down on me and my SD, as they know there will be no letting up.
I killed a 3700 Clawhammer about a year ago when I was insisting that it bench 3D@ 3ghz on water. As I recall, I was pumping 2 volts even thru it at the time.
Ya, when my old dfi nf-3 screwed up and gave 2.3v to the cpu, It posted I went into the bios saw that huge number and hit the switch on the back of the psu. But the cpu never posted again.
1.7v, 1.8v,1.9v and they were all phase change cooled to -60CQuote:
Originally Posted by Absolute_0
They died over the course of 1-2 Years
I now strongly suggest that you keep your Vcore at 1.55v or less when ever possible..
Oh and these chips were pushed to their max 24/7 to boost my folding score..
They just seemed to wear the :banana::banana::banana::banana: out..
other chips set at 1.6v and below seem to still be working nicely with no probs..
As I've always stated, high vcores are more damaging to your cpu's that high temps. The A64 may be able to withstand 2.0+Vcore for short periods of time, but you're definately cutting the life expectancy of the cpu way down. BTW, I can thank a faulty DFI for somehow killing one of the most awesome cpu's I've owned to date - an 0435 CLAW FX55 :(
just out of curiousity, when you guys report ur vcores, is it what you set in bios or what displays in windows?
for example, i have my vcore on a dfi at 1.45+110%
so when stating my vcore, would you state its:
1.45v
1.55v (what windows states it as)
or 1.45x1.1 = 1.595v?
and finally, assumings its 1.595 and running 24/7 at around 32-33c (full load of both cores, ambient when idle)...its not too high is it?
I fried a 3400 clawhammer. pumped 1.8 into it for months,one day wont post.
yup sure have it was a DTR3400 was trying to push 1.8v thru it with a v-dimm of 3.4v using a mach1 pasechange unit.... to this day the cpu sits on my desk in an ashtray to remind me that voltage kills......
peace perc,
mine was a mobile 62w on water. couldnt take the beating lol
figers cross never
the thing is
if you really want the mhz you just have to do it period
how could you just sit there and say are well thats all the oc i can get @ 1.55v and call your self an over clocker
and as for life expectance of the chip who here intends to have the same chip next year
I run Phase Change and I never exceed 1.6v. Folks we are talking a 90nm process here, electron migration is a reality and will be even more so once we go to 65nm process.
Excellent question. Run-time display is usually 0.07 volts lower than what I set in the BIOS on my SLI-DR.Quote:
Originally Posted by chinkgai
I burned my old wini 3200+ CBBFD when i pumped 1.95V in to him :D
Ok, you guys need to stop basing voltages off Windows readings. You need to compare your BIOS settings with a multimeter and then tell us hwo much your voltage really is set at... USually BIOS settings are quite correct for me (DFI LP NF4 Ultra-D)
Not yet! :D
But I burned a Barton two years ago! :D
I think A64s are stronger! :toast:
Quote of the day?Quote:
Originally Posted by bachus_anonym
So I guess I can't call myself an overclocker because I don't want to take the risk and kill a cpu which costs $600+ or one which is a pre-release sample and won't even be available for a few months....... :rolleyes:Quote:
Originally Posted by bazx
Well I like to keep most of my chips alive because when they stop being in my main rig, I put them to folding.. and when They die.. I sell them on Ebay
sledgehammer for 4 months, 1.775v 24/7. first time i mounted the hsf wrong and it went to 90c in bios, still running fine.
I think OPB take the crown here, he killed more cpu with high vcore than anybody else!! :D :D :D
pre release chip,sample,over volted 1.85vhttp://www.net7.info/1/sm/tongue.gifQuote:
Originally Posted by s7e9h3n
http://www.xtremesystems.org/forums/...495#post764495
http://www.net7.info/1/ok.jpg
http://www.net7.info/1/sm/yeah.gif
Ummm, if you didn't notice, I was being sarcastic and BTW, that quote is taken out of context, IIRC. I was speaking of a screenshot and not of sustained use @ that Vcore. Ok, Mr. OverclockingMasterGuru - please tell me you're running your FX57 @ 1.8+ Vcore for 24/7 use cuz if you're not, then you're not an OC'r :rolleyes:Quote:
Originally Posted by bazx
Quote:
Originally Posted by s7e9h3n
You need to cool down dude its only dots on a screen
ps: i like this word "OverclockingMasterGuru" can i use it it my sig? :D
I didn't think that you can check vcore with a DMM... if so, how?Quote:
Originally Posted by dmo580
your joking I hope...................................Quote:
Originally Posted by nn_step
Sure you can, I do it every time. Somebody posted the photo with the points you go to just recently.Quote:
Originally Posted by g0dM@n
And yes the OS readings from the sensors are off on my SLI-DR.
Here is what I got for two runs:
Code:
Volts 1:
- set: 1.325 * 110% = 1.4575
- reported OS: 1.41
- measured voltmeter: 1.468
Volts 2:
- set: 1.45 * auto = 1.45
- reported OS: 1.41
- measured voltmeter: 1.449
I am, In Reality I put them on poster board like trophys to my ablity to push stuff too far...Quote:
Originally Posted by racinjimy
Dont scale up that vcore, if you dont have the collin :)
I ran FX57 with -90C average evap. temp. for the past 2.5 months for about 7 hours daily @ 1.6/1.7V for surfing etc, and 1.8V for benching. Last week the HSF came off by itself and the cpu went topless. :)
LOL....OPB was telling me about that.....so it still works?Quote:
Originally Posted by Dumo
Mmm... naked A64...
I don't think I can kill mine using this K8N neo2 plat... max vcore on this bios is 1.65ish :(
you can do alot more on a Neo2 using corecenter (up to 1.9V).Quote:
Originally Posted by zenzog
RIP @4076Mhz :( ..Not because of Voltage...probly shorted or crushed core. The SLI-D and cascade still going strong though and colder too :)Quote:
Originally Posted by s7e9h3n
Time to hunt for another 0516 cabce...
PM me....I may know someone that can get you one ;)Quote:
Originally Posted by Dumo
I've had an 0441XPMW newcastle which loved voltage, i've pumped 2,05V through it and it even went faster than 1,75V/1,85V/1,95V on air. Just after one PiFast and SuperPi 1M run it shut down because it reached 70*C :p:
But CPU still going strong :)
Could you link me to where you found that info? I really thought it wasn't possible b/c the voltage goes right to the core. I figured if there were a spot that it would still be off, but I'll give it a shot if you link me. :)Quote:
Originally Posted by Martin Cracauer
One athlon xp-m @ 2.3V on air :D Well, actually NF7 undervolts a bit so it was like 2.25V. I knew I was gonna kill it because I was pissed off. :D Worked fine @2.1V though.
I forgot where I got it from, this is my local copy of the pic. I hope the original author doesn't, otherwise let me know.Quote:
Originally Posted by g0dM@n
http://wavehh.dyndns.org/tmp/dfi-vol...repoint3nw.JPG
yep killed my 3000+ venice with 1.9v and fan stoped working while I was at school and when i got home no more workey but the venice i got back from rma is nice also.
Wouldn't your mobo shut it down itself on an overheat?Quote:
Originally Posted by Mr_Slinky
It might, but I've gotten a winchester to hit 67C and keep loading at that temp. I did that when I first got my A64 a long time ago and wanted to OC it on stock cooling. If his fan stopped working then maybe the temp didn't go up so fast.Quote:
Originally Posted by Daved+
Think about it... there is a reason why it is called "load temp." That is where your cooling system pretty much bottoms out (or tops out actually) and stays at that constant temp. It all depends on your case cooling and the CPU cooling method. Think of falling speed due to gravity in reference to terminal velocity... there is a maximum.
What probably happened in his chip is that when the fan stopped working the chip went up in temp really fast. It probably eventually topped out at a certain point, let's say 80C. Then what happened is that the case airflow came into effect and created this new "maximum" cpu temp for his system. The system couldn't break 80C b/c it reached its critical point. If the mobo was set to shut off at a higher temp, then the chip probably maintained 80C for a long enough time to kill itself.
I'm not sure if this is what happened, but I'm just taking a wild guess. That, and also this is just generally speaking. If a CPU fan can cause a critical point for a LOAD TEMP, then it would only make sense to say that without a CPU fan the case fans will cause a new, yet higher, LOAD TEMP.