long term overclocking

K

kharan5876

Limp Gawd
Joined
Oct 3, 2007
Messages
141
I wanted to play with overclocking on my HT / Gaming PC.

It's a Core i5 750, P55 mobo, 8gb ddr3 1600.

I've always avoided overclocking because I was worried it would shorten the lifetime of my hardware.

I just purchased this i5 setup and the motherboard I was using was an asus P7P55D LE. I was playing with some overclock settings. I tried BCLK all the way up to 200 but it ran too hot. I eventually was using BCLK of 160 for the time being to get the memory up to 1600Mhz and the cpu to 3.2Ghz. After a few days the motherboard died. I'm willing to put the blame on there being a defect in the motherboard.

Anyway, I had a couple questions for you overclockers.

How long have you had a stable system that was overclocked throughout its whole lifetime? How "extreme" was this overclocking? Just a little bit over or as far as it could be pushed to be stable?

What are your thoughts and opinions on the affect of overclocking on hardware lifetime.
 
kharan5876 said:
I wanted to play with overclocking on my HT / Gaming PC.

It's a Core i5 750, P55 mobo, 8gb ddr3 1600.

I've always avoided overclocking because I was worried it would shorten the lifetime of my hardware.

I just purchased this i5 setup and the motherboard I was using was an asus P7P55D LE. I was playing with some overclock settings. I tried BCLK all the way up to 200 but it ran too hot. I eventually was using BCLK of 160 for the time being to get the memory up to 1600Mhz and the cpu to 3.2Ghz. After a few days the motherboard died. I'm willing to put the blame on there being a defect in the motherboard.

Anyway, I had a couple questions for you overclockers.

How long have you had a stable system that was overclocked throughout its whole lifetime? How "extreme" was this overclocking? Just a little bit over or as far as it could be pushed to be stable?

What are your thoughts and opinions on the affect of overclocking on hardware lifetime.
Click to expand...

I currently have a C2D E7200 that I got 1.5 years ago. It ran at 3.8ghz up until 2 months ago when it became unstable and I was unwilling to raise the voltages to maintain it. Now it runs at 3.6ghz with much less volts than before.
Keep in mind that at 3.8ghz it ran right at the upper limit of intel's specifications for wolfdale chips. I suspect it will run just fine at the current 3.6ghz forever.

But the sign of it dying has prompted me to purchase a core i7 920 system that arrives today ^_^
 
kharan5876 said:
I've always avoided overclocking...

I'm willing to put the blame on there being a defect in the motherboard.
Click to expand...

But not willing to put the blame on user error by someone who has never overclocked before? You could have easily put the wrong volts on something, or perhaps you overstrained your power supply and it damaged the board. Motherboards can also be just plain crappy. I had two MSI nforce boards (same type) that were fine overclocking an E6400 and a Q6600, but they both died several times from the higher FSB needed to push an E8400 to 3.8 (which isn't even extreme for that chip...mine was capable of 4.2). Were those boards defective? It's hard to say. I doubt they would have died if left at stock. Yet I don't think I was unreasonable in expecting a moderate overclock out of them. I chalk it up to them being a trash chipset (nforce) from a trash company (MSI).

If you buy good stuff, it should last a long time as long as you're not too extreme and don't go for your max OC. Stop before it starts taking a lot of extra volts to get the next 100mhz. Do that and there should be no troubles having it work 2-3 years, after which you sell it to someone who doesn't really care about OCing and it should function for a long time to come at stock speeds. The worst that should happen over those 2-3 years is that you occasionally have to bump the vcore a notch or two to keep it stable.
 
No it's very possible I could have screwed it up somehow but I didn't change the voltage very much at all. I started pushing it up in small increments to try to get the 200 BCLK's stable until the heat levels were too high and then stopped. I believe 1.35 was about the highest I went to before giving up.
 
Been running my Q6600 at 3.6Ghz with 1.488v(idle) and 1.41v(load) for 2 years. Not a hiccup.

Sold an Opty 165 and Asus A8V Deluxe to a friend. I had the chip clocked to 2.9Ghz for almost 2 years and he's running it at 3.0 for the past 2 years without a hiccup.

Don't think I've ever had to bump up voltage after a period of time to keep things stable. If that were the case I'd be looking at the motherboard or PSU for issues before even thinking it was the processor.

Keep it cool and keep the voltage under spec I don't see why a setup shouldn't last you the lifetime of its usefulness.
 
I've been running my E8400 at 3.8Ghz for over 18 months just fine. As long as you don't go stupid with the voltage they'll last longer than your desire to keep them.
 
Yep started OCing with a Tbird 1.4 and never looked back.
Do your homework before going crazy with voltages and get quality coolers since you normally reused them on and on and you will always get the most out of your gear.
Welcome to the addiction of OCing!
 
I've had my Athlon 64 3000+ at 2.412GHz (stock clock is 1.8GHz, so >30% overclock) for five years now. No instability, no issues.

Increases in voltage directly shorten expected life span based on the voltage increase. I don't think increasing the clock while maintaining the voltage directly shortens chip life span per se but the added heat might cause problems with insufficient cooling to compensate.
 
Err 3.2Ghz on an i5 should be easy. I'm at 3.5Ghz, at stock voltage (which is 1.17 during loads). I'm wondering if you increased VTT or vDIMM out of spec, since vDIMM in particular will harm both the RAM and the IMC.

You mention having 8GB - I presume you had 4 sticks of 2GB DIMMs correct? I know that some users had problems getting even a small OC when all slots were filled, but the moment they dropped down back to 4GB, it was a breeze.
 
I have never heard or experienced long term OCing as being a threat to the life of my or anyones machine

I had a P4 3.2 Prescott @ 4.2 on water for over a year (say closer to 2 years but it was a long time ago) without a problem.

I think you will be safe
 
Everyone likes to say OC'ing "shortens" the life span of the cpu. But there is never a lot of discussion on what exactly is the lifespan of a cpu.

For the majority of the readers here at Hardocp, we will be upgrading before we ever run to the issue of our hardware dying before it was meant to due to oc'ing. (Assuming ample cooling, not extreme LN2 Cooling, dunking it into mineral water, etc)

The biggest liabilities as far as I am concerned is the capacitors on the motherboard/psu. Those are the things that I see go bad the most computers over several years of usage.

I have been running my system in my sig for just about 6 months now OC'ed and I don't plan on changing the OC till I get rid of the computer.

Extreme Heat and Extreme Voltage will kill stuff, if you are careful the risks are greatly diminished.
 
I think I read somewhere that Intel, on average through tests, found the lifespan of a CPU to be roughly 30 years under normal operating conditions.
 
Stay under the max Vcc and Vtt listed on Intel's processor datasheets and you'll be fine (which is roughly .05 volts above VID spec).

I've had 65nm duals at or just under VID spec for 18 months, and now have a 45nm quad (Q9450) above VID but under max via datasheet for 12months+; no problems on any of them.

VID spec and max Vcc and Vtt are there for a reason; go above them and yes you risk frying your $#!t. I would recommend you stay UNDER Vtt max; AT fried a QX9650 w/ 1.45 Vtt, even though that's the max listed...I personally stay under 1.4 Vtt (and that's what AT recommends people do). The new 45nm chips are less tolerable of high Vtt than the 65nm ones were.

(Obviously voltage is only one piece of the puzzle - cooling is the other.)
 
vengence said:
I've been running my E8400 at 3.8Ghz for over 18 months just fine. As long as you don't go stupid with the voltage they'll last longer than your desire to keep them.
Click to expand...

I went stupid and has not shown any signs of slowing down on my E5200 for over 14 months, Still strong at 4.05ghz @ 1.45V.
 
Casper312 said:
I went stupid and has not shown any signs of slowing down on my E5200 for over 14 months, Still strong at 4.05ghz @ 1.45V.
Click to expand...

I think I'm at 1.38command 1.35V actual or something. I'm limited by not being able to lower the GPLVREF on my 780i, oh well would probably only have given me a few hundred mhz more anyways. :(
 
Mav451 said:
Err 3.2Ghz on an i5 should be easy. I'm at 3.5Ghz, at stock voltage (which is 1.17 during loads). I'm wondering if you increased VTT or vDIMM out of spec, since vDIMM in particular will harm both the RAM and the IMC.

You mention having 8GB - I presume you had 4 sticks of 2GB DIMMs correct? I know that some users had problems getting even a small OC when all slots were filled, but the moment they dropped down back to 4GB, it was a breeze.
Click to expand...

Yes its 4x2gb
I kept the ram voltage at 1.7. The memory is rated at 1.65 but I could only change it in .1 increments.
 
If you're really interested in the topic of temps and voltages, you might want to read this. I've learned a lot from the thread.
 
I had a Pentium 4 2.4C Northwood at 3.0 Ghz for at least 3 or 4 years.
 
JCDenton said:
But not willing to put the blame on user error by someone who has never overclocked before? You could have easily put the wrong volts on something, or perhaps you overstrained your power supply and it damaged the board. Motherboards can also be just plain crappy. I had two MSI nforce boards (same type) that were fine overclocking an E6400 and a Q6600, but they both died several times from the higher FSB needed to push an E8400 to 3.8 (which isn't even extreme for that chip...mine was capable of 4.2). Were those boards defective? It's hard to say. I doubt they would have died if left at stock. Yet I don't think I was unreasonable in expecting a moderate overclock out of them. I chalk it up to them being a trash chipset (nforce) from a trash company (MSI).

If you buy good stuff, it should last a long time as long as you're not too extreme and don't go for your max OC. Stop before it starts taking a lot of extra volts to get the next 100mhz. Do that and there should be no troubles having it work 2-3 years, after which you sell it to someone who doesn't really care about OCing and it should function for a long time to come at stock speeds. The worst that should happen over those 2-3 years is that you occasionally have to bump the vcore a notch or two to keep it stable.
Click to expand...

HA, a P7N Diamond I take it? I killed one too with an e8400 C0 at 425x9. That northbridge ran hot.

Anyways as long as you are willing to read up on what you are doing you should be fine. Most problems are caused by people overclocking while leaving everything on Auto which tends to overvolt most settings on most boards.
 
My old Athlon X2 4400 has been OC'ed since the day I got it. It's still makes a pretty decent backup/media server. The only times it wasn't OC'ed was during the original install and a BIOS update a few years later :D
 
Have been running my Q9650 at 4.2 - 4.4Ghz for about 10 months now 24/7.

Haven't seen any signs of degradations and I run stress tests weekly with old and tested out settings. If anything I've seen slight signs of the chip needing less voltage but as any other 45nm Yorkfield its finicky about ambient temp.
 
You must log in or register to reply here.
Back
Top