How long are your overclocked amd pocessors lasting?

C

Ciggarilo Himself

Limp Gawd
Joined
Nov 22, 2006
Messages
464
How long are your guys overclocked amd processors lasting.

Please include:how long your processor has been running at this speed,voltage and your temps.

Also what do you guys condsider a safe voltage for 90nm amd chips?

Thanks;)
 
I had a 3200+ @ like 2.6GHZ at 1.57v with a XP90 for a year without any probs. I have it in a "server" running 24/7 at default now.
 
lol, IMO, I don't think anyone who overclocked wisely
out lasted their CPU.

It is most likely you'll change your comp before you proc
goes dead :rolleyes:

But any ways, I had a 4400+ X2 (s939) for a year and a half,
OC'ed @ 2.6ghz, air cooled and ran that comp for around
24/5 each week, and still running strong.

It isn't with me right now; back at my parent's place,
but I think it was on 1.45~1.47V.

Temp, I can't remember, but it wasn't at a range I worried
about, or else I would have remembered it :p

Basically to sum it up...

CPU: AMD64 X2 4400+ (s939)
OC: 2.6GHz (260x10) from stock 2.2GHz
Voltage: 1.45~1.47v
Cooling: Air (Tt big typhoon)
Temp: Can't remember
Usage: Daily, moderate gaming and multimedia purposes
Duration: Approximately 1.5 years
 
3000+ Winchester @ 2.45 for about 2 years @ 1.55volts with Thermaltake Volcano. Average temp is about 58C because I run Folding@Home almost 24/7. As one of the posters said most people will get a new proccessor before their proccessor dies from OCing. I'm planning on getting a 165 soon too, but this proccessor has served me well.
 
3800+ @ 2.7, half its life at 1.5, other half at 1.55 (dif motherboard that required more voltage to get same speed). 1 year old and going strong.
 
Opty 144 somewhere between 2.4 and 2.7 depending on the season. @ 1.55v for 2.7 it idles at 31 and loads to 45. Been running for over a year now with no problems, unless you count when I change the HT speeds without adjusting the mem divider and try to run my pc3200 @ 230 2-3-2-6-1T ;)
 
Hmm...Good feedback I was just curious if runnig at higher voltages lessened life by alot,Looks like I have nothing to realy worry about.
 
In general,
if you know what you are doing and have sufficient
cooling, there is not reason for your processor to
be significantly effected by overclocking :D
 
Riddlinkidstoner said:
1 month tomorrow.

Opteron 165 @ 3ghz. 1.4v 24c idle 34 load on AC Freezer 64 Pro.
Click to expand...
Wow, how are you getting these ridiculously low temps? I have the same CPU, same heatsink, running a lower voltage, and my temp delta is 20-22C between idle and load.

To OP, had a A64 Winchester 3000+ @ 2.4Ghz/1.45V for 2 years. Recently sold it for a Opty 165.
 
Yeah, I was wondering about these low temps on air. As far as I can tell my Big Typhoon is utter crap, my 3400+ Venice (2.7GHz @ 1.6v) hovers between 38 and 39c while idle, and load temps aren't all that stunning either...maybe one of my heat pipes isn't piping heat.

In any case, the rule of thumb is a 10c jump in temps usually halves the lifespan of your processor. If you were to, say, pump up the voltage and you have your processor running 20c higher across the board, the processor will generally only last 1/4 as long as it would have at stock temps (which is still a ridiculously long lifespan btw). As long as your temps aren't crazy high you relay don't have anything to worry about.
 
16 months now on my overclocked X2 @2.6ghz. And it still is kicking some ass at high res and iq settings, along with my 8800GTS.


June, I will be upgrading mobo/cpu/ram again. Pre quakecon upgrade :p
 
processor in the computer im typing on has been using an Opteron 165 running at 2.97ghz/1.475 volts for only 2 weeks now. :)

my other rig has has an Opteron 165 running at 2.90ghz/1.375 volts...its been at that speed for 6 or 7 months now?

i had an opteron 170 o/ced to 2.85ghz/1.45 volts running for a year.

also sold an opteron 170 to my bro which is running at 2.5ghz/1.425 volts, also a year old.
my brother has sold it to someone else now...no issues.

24/7 on all of these chips(except for hardware changes:)). i run folding@home on my machines.
 
Unknown-One said:
Yeah, I was wondering about these low temps on air. As far as I can tell my Big Typhoon is utter crap, my 3400+ Venice (2.7GHz @ 1.6v) hovers between 38 and 39c while idle, and load temps aren't all that stunning either...maybe one of my heat pipes isn't piping heat.

.
Click to expand...


Most of those low temps are erroneous readings......unless you live in a truly cold climate, have water GOOD cooling, something like that. If you ambient temp is say, 27c, there is no way in hell your cpu is going to load up to only 34c under load.


What are your load temps on that cpu? My X2 runs at 52c to 54c under load according to CoreTemp. BIOS and MBM5 report almost 10c lower.....that's at 2.6ghz and 1.45 vcore.
 
Every one knows high temps are bad,But WHY are high temps actualy bad,does anyone here actualy know WHY they are bad,I presume it is just because the heat makes the transistors expand more thus wearing them out faster maybe.
 
I use a probe mounted on the side of the ihs on my 3200 venice. Its a compunurse temp readout.my Idle temp is 30-32c depending on ambient here in orlando and load temps with prime is 39-42c. This is with voltage of 1.37v on the cpu. I have found the compunurse probe to be very accurate and as crazy as it seems it matches up with motherboard monitor exactly on my asus a8n-e. This is the only board ive run in the last few years that has temps match up. Most of the time the temps are hotter than the software is reporting. If you that concerned with temps purchase a couple of compunurse probes[$8.00 ea] and put them in front of your window on your case and youll see them all the while your rig is running. I monitor cpu and nf4 chip with these probes.
 
Ciggarilo Himself said:
Every one knows high temps are bad,But WHY are high temps actualy bad,does anyone here actualy know WHY they are bad,I presume it is just because the heat makes the transistors expand more thus wearing them out faster maybe.
Click to expand...
Not quite. There are several reasons why high temperatures are undesirable for a CMOS circuit:


1. Higher Heat = Higher Resistance. Consider a MOSFET (Metal Oxide Semiconductor Field Effect Transistor) inside your CPU. The Gate Capacitance and the Channel Resistance in series effectively forms a 1st order low pass filter, whose time constant (tau) = R*C. The cut-off frequency of this (unintentional) low pass filter is 1/tau, or 1/(R*C), since frequency = 1/time_period. This frequency is the maximum frequency the MOSFET can switch at. Obviously, increase the R (resistance) causes the the cutoff frequency to decrease. Which leads to lower maximum operating frequency of your processor, thus limiting your OC. But this is only one part of the equation, read on...


2. Signal propagation time increases as temperature rises, as signals travel slower in hotter temperature. An over simplified model of the time delay as signal travels from one flip-flop (basic circuit component) to the next can be constructed as follows:

Delay_total = Gate_delay + Propagation_delay + Setup_time + Strobe_time + Hold_time + Clock_skew

We won't worry about the last 4 items for now. Gate_delay is explained in #1 already. Propagation_delay is the time it takes to traverse through the wire that connects two flip-flops. The minimum time period is equal to the Delay_total. An increase to any of the delays in the equation raises the total delay, which lowers the max operating frequency (again, frequency = 1/time_period) . So again higher temps make your OC suffer.


3. This one is pretty simple. If temperature gets too high, it can physically destroy the circuit.


4. If heat doesn't kill your CPU outright, prolonged high temperature exposure can lower its reliability. This is caused by electromigration. In simple terms, the materials inside your circuit gradually move from their original positions, which can form gaps and cracks. This could eventually lead to the loss of one or more interconnects, and lead to the failure of the entire circuit.


Anyways, I'm not sure if I did a very good job explaining, feel free to ask any questions :p
 
morfinx said:
Not quite. There are several reasons why high temperatures are undesirable for a CMOS circuit:


1. Higher Heat = Higher Resistance. Consider a MOSFET (Metal Oxide Semiconductor Field Effect Transistor) inside your CPU. The Gate Capacitance and the Channel Resistance in series effectively forms a 1st order low pass filter, whose time constant (tau) = R*C. The cut-off frequency of this (unintentional) low pass filter is 1/tau, or 1/(R*C), since frequency = 1/time_period. This frequency is the maximum frequency the MOSFET can switch at. Obviously, increase the R (resistance) causes the the cutoff frequency to decrease. Which leads to lower maximum operating frequency of your processor, thus limiting your OC. But this is only one part of the equation, read on...


2. Signal propagation time increases as temperature rises, as signals travel slower in hotter temperature. An over simplified model of the time delay as signal travels from one flip-flop (basic circuit component) to the next can be constructed as follows:

Delay_total = Gate_delay + Propagation_delay + Setup_time + Strobe_time + Hold_time + Clock_skew

We won't worry about the last 4 items for now. Gate_delay is explained in #1 already. Propagation_delay is the time it takes to traverse through the wire that connects two flip-flops. The minimum time period is equal to the Delay_total. An increase to any of the delays in the equation raises the total delay, which lowers the max operating frequency (again, frequency = 1/time_period) . So again higher temps make your OC suffer.


3. This one is pretty simple. If temperature gets too high, it can physically destroy the circuit.


4. If heat doesn't kill your CPU outright, prolonged high temperature exposure can lower its reliability. This is caused by electromigration. In simple terms, the materials inside your circuit gradually move from their original positions, which can form gaps and cracks. This could eventually lead to the loss of one or more interconnects, and lead to the failure of the entire circuit.


Anyways, I'm not sure if I did a very good job explaining, feel free to ask any questions :p
Click to expand...


Very good explanation clarified it alot for me,I thought electromigration only occured from higher voltages...guess I was wrong.
 
Ciggarilo Himself said:
How long are your guys overclocked amd processors lasting.

Please include:how long your processor has been running at this speed,voltage and your temps.

Also what do yo guys condsider a sfe voltage for 90nm amd chips?

Thanks;)
Click to expand...

I dont remember vcore/temps, all amd processor except my first amd "850":mad: was great overclock cpus. Before my amd tour I o.c:ed a celeron 300a to 450, very stable.

Amd;
*850 F0P38 :eek:
*XP2500 Barton @XP3200 *sold*
*XP2400 Mobile @XP3200 -> Could o.c more but it wasn't necessary *sold*
*XP1700 Thouroughbred @ 2,3Ghz Using as fileserver etc.

I have run all my setups 24/7 and my xp1700 has been running for some years now :D
The cpu cooling on this baby is a Thermalright SLK-947U with a papst 120mm.
Other cpu coolers i've tried ; Some Arctic Cooling, Zalman CNPS7000A-ALCU, Thermalright SLK-947U
Swiftech MCX462-V, PAL8045T "Alpha"
 
air cooled 3700+ @ 2.55ghz w/ 1.49v

its been goin well for about 1.25 yrs so far, with load temps maxing at about 45*C w/ my xp-120
 
Ciggarilo Himself said:
Very good explanation clarified it alot for me,I thought electromigration only occured from higher voltages...guess I was wrong.
Click to expand...
Both increased current density (higher voltage) and higher temperature causes electromigration. It happens during a phenomenon called momentum transfer, basically moving electrons collide with the ions in your circuit, and causes them move from their original position. Crystal lattices vibrate as temperature goes up (thermal vibration), which makes it much more likely the an electron would collide with a defect site.

Anyways, if you interested in learning more and have some background in semiconductor physics and device fabrication, Wikipedia has a GREAT article on it :)
 
morfinx said:
Both increased current density (higher voltage) and higher temperature causes electromigration. It happens during a phenomenon called moment transfer, basically moving electrons collide with the ions in your circuit, and causes them move from their original position. Crystal lattices vibrate as temperature goes up (thermal vibration), which makes it much more likely the an electron would collide with a defect site.

Anyways, if you interested in learning more and have some background in semiconductor physics and device fabrication, Wikipedia has a GREAT article on it :)
Click to expand...

So is electromigration what caused the Northwoods to die early when they were using higher voltages(S.D.N.S sudden northwood death syndrome).
 
Server: 754 3000+ Vinice, 1.65v, 2.75Ghz, 40/60C, since 1/11/2006 has been alive as my download Server for this long and going.

Main Box: AM2 X2 3800+, 2.9ghz, 1.55v, 35/65C (Reason small case.) since 6/21/2006 so been running like this for about 6 months.
 
AMD Athlon 64 3500+ 2.2GHz @ 2.55GHz (255x10 @ 1.500v)
Idle: 34-36c (Not positive, I run UD 24/7, been a long time since idle)
Load: 48-52c (Depends on room temp :p)

Been running like that for about a year, UD has been running since... hmm, September.
For about 4 weeks, it was pushed to 2.706GHz, loading to 58c :)eek:).

Oh, air cooled, stock cooler..
 
Man some of these overclocks are awesome,Would I be correct to say that AM2platform is much better for overclocking than s939?;)
 
I'm running dual 35 watt 2400+ AMD XP-Mobile Bartons, both overclocked to 2.4GHz and pumped with 1.75v. Idle temps run around 45c to 50c, and load is around 60c (Vantec Aeroflows and not a lot of space for two larger heatsinks on the board). At times it'll reach 65c, and reboot...but it has been running strong for about 3 years now...and even survived a spectacular powersupply failure...
 
Tim_axe said:
I'm running dual 35 watt 2400+ AMD XP-Mobile Bartons, both overclocked to 2.4GHz and pumped with 1.75v. Idle temps run around 45c to 50c, and load is around 60c (Vantec Aeroflows and not a lot of space for two larger heatsinks on the board). At times it'll reach 65c, and reboot...but it has been running strong for about 3 years now...and even survived a spectacular powersupply failure...
Click to expand...

Are these in the same machine?
 
Processor in My Sig Running Strong since April. Ive been running it at 2.91 (3.23*9) at 1.425 since then. Temps never exceed 45c
 
i've had a barton since.. late 2002, a 3200+ clawhammer that got MUCH abuse that I am typing with right now, got it back in mid '04

here's what it looks like ;)

core.jpg
 
2 old overclockers still working :

Celeron 300A to 464 MHz, still working today :) It's now retired.

AMD thunderbird 1200 @ 1450 MHz, still chugging happily in my dad's PC.

Current computers :

NewCastle 3000+ @ 2.4 GHz still working in my file/media server without issues, folding 24/7 as well.

Venice 3000+ @ 2.5 GHz, folding 24/7 ;)

 
s754 3700+ Clawhammer @1.56v for 2 years now. Rock solid.
s754 3400+ Venice @ 1.62-1.66v for 6 months (see sig). Home gaming/personal productivity system. Steady and solid.
Playing with a new s754 3200+ Venice for the last couple of days. Still sorting it out. I've had it up to 256x11 @1.66-1.7v trying to find a stable oc but no luck so far. It will run Superpi and internet/Windows apps ok but chokes while gaming.
It is, however, very stable at 236x11 @ 1.48v and runs in the mid-upper 40C range under Prime95.
 
Had a venice 3200 @ 2.6 (1.5v) for 6 months, thats now in my sister's rig.
I now have an opty 165 @ 2.5 (1.5v, idle low 30's, load high 40's z9700 @ lowest fan) I've got at christmas .
 
A64 3700+ @ 2.77GHz, 1.6v - ran for over a year with no problems (until I replaced it with an opteron) folded 24/7 for about 6 of those, stock cooling, load temp about 50-55c.

My new opty 175 is at 2.6ghz, 1.55v (really shit stepping I guess) and both cores load temp about 40-43c with a Tuniq Tower
 
opteron 165 @ 2.9ghz and 1.45v 28C idle and 39C load (speedfan) running 24/7 for about 6 months now
 
You must log in or register to reply here.
Back
Top