Need some Overclocking Advise

[Jon Lai]

Believe it or not, this is actually my first time overclocking. My system:

P4 2.8 Northwood HT (200Mhz x 14 multiplier)
Asus P4P8X SE
OCZ 2x 512mb Premier 2.5-3-3-7

Right now, I've OCed to 230Mhz FSB, my specs now are as follows

P4 running at 3220Mhz
2x 512mb Premier running at 2.5-4-4-8

Do you guys think I'll be able to go much higher than 3220Mhz? What do you guys think will be my max limit with the hardware I have?

Also, what will be the device that will limit my OC - the processor, mobo, or RAM?

BTW, I've never seen anyone go higher than 3-4-4-8 for overclocking. Since my Premier set is already running at 2.5-4-4-8 does that in anyway hint that I won't be able to go much farther?

[Amourek]

Your CPU should be able to hit 230Mhz no problem. Your memory on the otherhand probably doesn't overclock well (Premier = Value = Budget) so you will most likely have to run it slower than the CPU... probably a 5:4 ratio which means 230Mhz CPU/184Mhz RAM - technically your RAM will be underclocked so you might be able to get the timings down.

I'd be most worried about your mobo - the P4P8X SE uses the 865P (inferior to the 865PE) chipset which if IRC is actually a modified version of the older E7205 Granite Bay chipset (which could only handle 133Mhz officially). So running your 2.8Ghz at stock speed is actually a 50% overclock for the mobo. I have no idea if it can handle another 30Mhz.

[tdotcbc84]

wow ~! i didnt know that ! very insightful dude!

[Jon Lai]

Your CPU should be able to hit 230Mhz no problem. Your memory on the otherhand probably doesn't overclock well (Premier = Value = Budget) so you will most likely have to run it slower than the CPU... probably a 5:4 ratio which means 230Mhz CPU/184Mhz RAM - technically your RAM will be underclocked so you might be able to get the timings down.

I'd be most worried about your mobo - the P4P8X SE uses the 865P (inferior to the 865PE) chipset which if IRC is actually a modified version of the older E7205 Granite Bay chipset (which could only handle 133Mhz officially). So running your 2.8Ghz at stock speed is actually a 50% overclock for the mobo. I have no idea if it can handle another 30Mhz.

Sorry if my post was confusing, but as I'm typing this, my CPU is already running at 3.22Ghz, so perhaps you didn't recall correctly, as if you did, I shouldn't be able to boot up Windows at this speed. Right now, both my CPU and RAM are running at 230Mhz, give or take a few Hertz or two.

What I want to know is if it can go even higher than just an extra 30Mhz clock.

BTW, according to CPU-Z, the chipset is Intel i865P/PE/G/i848P, Rev. A2. Are you perhaps suggesting that running 3.22Ghz right now is dangerous for my CPU, that I should immediately downclock it ASAP?

[Jon Lai]

bump

help someone?

[goofball]

What I want to know is if it can go even higher than just an extra 30Mhz clock.

BTW, according to CPU-Z, the chipset is Intel i865P/PE/G/i848P, Rev. A2. Are you perhaps suggesting that running 3.22Ghz right now is dangerous for my CPU, that I should immediately downclock it ASAP?

Nobody can give you a definite answer. What cooling are you using? What voltage are you at now for the CPU? How hot is the northbridge chipset/heatsink getting? Those are definitely variables which can affect your ability to overclock.

The 865P is not the same as the 865PE in terms of overclockability. What Amourek (and I am agreeing with) is that it is inferior to the 865PE and built off an older chipset which didn't officially support 800FSB but 533. So potentially, the chipset was already being overclocked to support 800FSB, and pushing it further may be an issue. You've already gotten it pretty far, IMO. 2.8C's tend to average 3.4-3.6Ghz, though the latter is with some high voltage.

[Jon Lai]

Nobody can give you a definite answer. What cooling are you using? What voltage are you at now for the CPU? How hot is the northbridge chipset/heatsink getting? Those are definitely variables which can affect your ability to overclock.

The 865P is not the same as the 865PE in terms of overclockability. What Amourek (and I am agreeing with) is that it is inferior to the 865PE and built off an older chipset which didn't officially support 800FSB but 533. So potentially, the chipset was already being overclocked to support 800FSB, and pushing it further may be an issue. You've already gotten it pretty far, IMO. 2.8C's tend to average 3.4-3.6Ghz, though the latter is with some high voltage.

Cooling - none, just fan + heatsink for the CPU, plus an exhaust 80mm case fan
Voltage - Don't know, haven't changed from stock
Temp - Was at 35C idle BEFORE I OCed, I'll check tomorrow for the new temp

How do you check the voltage that your CPU and memory is running at? I don't seem to be able to find it in BOIS nor from CPU-Z. Perhaps another 3rd party software?

I kinda understand what you guys are saying now. I actually bought this system prebuilt, and it came bundled like this, so I'm actually surprised to hear that the guy perhaps didn't give me a perfect fit for my system? And I gave him $1380's worth of business!

So, would you suggest that running 3.22Ghz (230Mhz) be an issue for my situation? Should I change it back to stock speeds for the potential health of my CPU/mobo?

Again, I'm an OCing n00b, but when you said "high voltages" with 3.4-3.6Ghz clocks, I persume "high voltages" = impact on CPU lifespan, would that be correct?

Should I be satisfied at 3.22Ghz?

[goofball]

How do you check the voltage that your CPU and memory is running at? I don't seem to be able to find it in BOIS nor from CPU-Z. Perhaps another 3rd party software?

So, would you suggest that running 3.22Ghz (230Mhz) be an issue for my situation? Should I change it back to stock speeds for the potential health of my CPU/mobo?

Again, I'm an OCing n00b, but when you said "high voltages" with 3.4-3.6Ghz clocks, I persume "high voltages" = impact on CPU lifespan, would that be correct?

Should I be satisfied at 3.22Ghz?

You could try the Asus Probe software. I'm sure it is supported by Asus Probe.

Not to sound like a goof but no one really cares about idle temperatures. Do you use your computer? Very rarely will a computer really show big problems at idle. It is load that you will see issues, unless the system is completely and hopelessly unstable, which is when you will see issues at idle.

The issue about buying full systems without prior hardware knowledge is that they usually just put things together that work at stock. Hardly anyone advocates overclocking, especially system builders. That's more problems that will come back to their store when things go wrong.

Measure all your temps if you can. If your setup is within range, you shouldn't worry.

Myself, I use my computer all the time, so I try to extract all the performance I can get. I dont' feel it is too fast for me, I feel it is actually too slow. Things don't happen right away, so it's still too slow.

Voltages can decrease the life but so can temperatures. Keeping them both at reasonable levels, you shouldn't have your hardware die on you before you would change them anyways. Can you imagine running a CPU/motherboard/ram combo that is 10 years old now? That would be the equivalent of a Pentium 166 with 16MB of EDO RAM!

[jdi_knght]

Again, I'm an OCing n00b, but when you said "high voltages" with 3.4-3.6Ghz clocks, I persume "high voltages" = impact on CPU lifespan, would that be correct?

Should I be satisfied at 3.22Ghz?

Do a google search for "sudden northwood death syndrome". I wouldn't go too high on the voltage, but other than that I don't forsee any major short-term issues.

I have a P4 2.8C on an asus P4P800, and around 3.2Ghz, things would start to get flaky. Others have gone up past 3.4 with ease, so YMMV.

Personally, I'd probably be satisfied with 3.22. Try running Prime95 for a day or two and see if it really is stable. If it is, you could try squeezing it up more a little bit at a time, and you'll probably eventually get to the point where you either have to up the voltage, play with some bios settings, and play with some memory settings to keep things stable. Just take it slow, watch the temps, and don't go crazy on the voltage.

Good luck!

[Jon Lai]

You could try the Asus Probe software. I'm sure it is supported by Asus Probe.

Not to sound like a goof but no one really cares about idle temperatures. Do you use your computer? Very rarely will a computer really show big problems at idle. It is load that you will see issues, unless the system is completely and hopelessly unstable, which is when you will see issues at idle.

The issue about buying full systems without prior hardware knowledge is that they usually just put things together that work at stock. Hardly anyone advocates overclocking, especially system builders. That's more problems that will come back to their store when things go wrong.

Measure all your temps if you can. If your setup is within range, you shouldn't worry.

Myself, I use my computer all the time, so I try to extract all the performance I can get. I dont' feel it is too fast for me, I feel it is actually too slow. Things don't happen right away, so it's still too slow.

Voltages can decrease the life but so can temperatures. Keeping them both at reasonable levels, you shouldn't have your hardware die on you before you would change them anyways. Can you imagine running a CPU/motherboard/ram combo that is 10 years old now? That would be the equivalent of a Pentium 166 with 16MB of EDO RAM!

Well that was because I didn't know how to, thanks for recommending Asusprobe to me.

Here is my monitor at default thresholds:

http://img490.imageshack.us/img490/7809/asusprobe19av.jpg

How am I doing? I haven't changed any settings to overclock except the FSB. I've heard about the sudden deaths of Northwoods and thus I'm not even going to touch the VCore settings - I'd rather run slower and keep my computer.

Am I "within range" as you said with these temperatures?

[Jon Lai]

I have a P4 2.8C on an asus P4P800, and around 3.2Ghz, things would start to get flaky. Others have gone up past 3.4 with ease, so YMMV.

Personally, I'd probably be satisfied with 3.22. Try running Prime95 for a day or two and see if it really is stable. If it is, you could try squeezing it up more a little bit at a time, and you'll probably eventually get to the point where you either have to up the voltage, play with some bios settings, and play with some memory settings to keep things stable. Just take it slow, watch the temps, and don't go crazy on the voltage.

Good luck!

Thanks for the personal experience, lol. How in your sense does "things start to get flaky"? Right now, after Amourek's remarks, I'm not as concerned about getting it even higher than 3.22Ghz, rather, I'm more interested to know if even STAYING at 3.22ghz is even safe for my system. Otherwise, if it isn't, I'd honestly rather go back to stock than risk it - I can't afford another computer anytime soon.

[goofball]

Well that was because I didn't know how to, thanks for recommending Asusprobe to me.

Here is my monitor at default thresholds:

http://img490.imageshack.us/img490/7809/asusprobe19av.jpg

How am I doing? I haven't changed any settings to overclock except the FSB. I've heard about the sudden deaths of Northwoods and thus I'm not even going to touch the VCore settings - I'd rather run slower and keep my computer.

Am I "within range" as you said with these temperatures?

Personally, that's about the upper limit of what I would call acceptable. I don't feel comfortable with 50+ c load on anything but prescott, and even then, 55c is all i feel comfortable with.

Like I stated before, make sure your other components are cool. RAM and Northbridge chipset (as well as mosfets) are kept within reason. If you can't keep your finger on the Northbridge chipset for 5 seconds without burning, that's too hot.

[Jon Lai]

Personally, that's about the upper limit of what I would call acceptable. I don't feel comfortable with 50+ c load on anything but prescott, and even then, 55c is all i feel comfortable with.

Like I stated before, make sure your other components are cool. RAM and Northbridge chipset (as well as mosfets) are kept within reason. If you can't keep your finger on the Northbridge chipset for 5 seconds without burning, that's too hot.

I heard from somewhere that the point of thumb is that anything lower than 70C on some overclocker's forums, and you said your comfortable level is 50C? Perhaps I should return to 2.8Ghz or maybe down to 3.0Ghz then, since that screen was taken when I was doing nothing except for running Bittorrent and MSN in the background...

BTW, by default, the threshold in PC Probe is 72C for CPU, and the mobo temp threshold is 75C! Are those numbers inflated or something?

And it seems like my mobo temp is much lower than my CPU temp - is that normal? Becuase it seems like my mobo/chipset is pretty cool, but my CPU temp is way off the scale. As I'm talking right now, I'm doing some CPU-intensive video conversions, and I'm reaching 60-66C... I feel unsafe now that you said 50C is your expectations.

[HUH]

That seems like a little too hot if 49C is the idle cpu temp. Also have you try running Prime95 and memtest? These programs will do stress testing on your system so you know if it's a stable oc.

[Jon Lai]

That seems like a little too hot if 49C is the idle cpu temp. Also have you try running Prime95 and memtest? These programs will do stress testing on your system so you know if it's a stable oc.

Right now my video conversion is finished and over 3 minutes it dropped back ot 49C, MB temp at 39C, so 49C does seem like the idle temp.

Keep in mind, though, I'm using the stock cooling - my computer was bought as a clone, so I've no idea how the guy installed everything. I'm thinking perhaps I should go back to 2.8Ghz or go down a notch to 3.0Ghz and see if the temps are acceptable first, since as you guys suggested, the temps are too hot even if the OC is stable anyways.

Or perhaps I need a new CPU cooler, but if so, I'd rather no OC, IMO, becuase I'm not an enthusaist anyways.

[goofball]

I heard from somewhere that the point of thumb is that anything lower than 70C on some overclocker's forums, and you said your comfortable level is 50C? Perhaps I should return to 2.8Ghz or maybe down to 3.0Ghz then, since that screen was taken when I was doing nothing except for running Bittorrent and MSN in the background...

BTW, by default, the threshold in PC Probe is 72C for CPU, and the mobo temp threshold is 75C! Are those numbers inflated or something?

And it seems like my mobo temp is much lower than my CPU temp - is that normal? Becuase it seems like my mobo/chipset is pretty cool, but my CPU temp is way off the scale. As I'm talking right now, I'm doing some CPU-intensive video conversions, and I'm reaching 60-66C... I feel unsafe now that you said 50C is your expectations.

70c for a Northwood is way too high, IMO. Too close to thermal specification for most 2.8C's, which is 75c.

Those numbers are definitely inflated. 75c motherboard temp, you'd probably get crashing or something worse happening. Mosfets tend to not do so well at such a high temp, and I can't say for sure what the ones that are installed are rated up to.

FWIW, I'm somewhat extreme but I have seen what happens when CPU's get too hot. The results are usually not too good. And the CPU temp usually affects the other components around it as well. So if the CPU temp goes that high, imagine the power regulators that are feeding it the power it needs. You can bet they're just as nice and toasty.

Stability doesn't necessarily mean longevity. The equivalent, i guess, to help you understand, would be to run at redline in your car....sure, you could do it for a certain amount of time and back off and it's all good but how often do you think you could do it before something bad happens?

The other thing to consider is that the onboard temp sensors are not accurate. That is completely possible.

[Jon Lai]

70c for a Northwood is way too high, IMO. Too close to thermal specification for most 2.8C's, which is 75c.

Those numbers are definitely inflated. 75c motherboard temp, you'd probably get crashing or something worse happening. Mosfets tend to not do so well at such a high temp, and I can't say for sure what the ones that are installed are rated up to.

FWIW, I'm somewhat extreme but I have seen what happens when CPU's get too hot. The results are usually not too good. And the CPU temp usually affects the other components around it as well. So if the CPU temp goes that high, imagine the power regulators that are feeding it the power it needs. You can bet they're just as nice and toasty.

Stability doesn't necessarily mean longevity. The equivalent, i guess, to help you understand, would be to run at redline in your car....sure, you could do it for a certain amount of time and back off and it's all good but how often do you think you could do it before something bad happens?

That's what I thought, it sounded ridiculous that the default threshold were so high.

BTW, what are mosfets?

And in case you don't know, I'm more for longevitiy instead of the faster processes for a given short time I'm using the CPU. I'm still running at 3.22Ghz, idleing right now. According to what I want (longevitity over OC), what speed do you suggest I run at?


The other thing to consider is that the onboard temp sensors are not accurate. That is completely possible.

So you're suggesting that the 60-66C temps while encoding video could possible be a problem with the temp sensors on the mobo?

If so, that makes it even harder to make sure where I should stay at the OC, or not to OC at all, even... What's your advice on this? Go back to 2.8Ghz, or try 3.0Ghz instead?

[goofball]

That's what I thought, it sounded ridiculous that the default threshold were so high.

BTW, what are mosfets?

And in case you don't know, I'm more for longevitiy instead of the faster processes for a given short time I'm using the CPU. I'm still running at 3.22Ghz, idleing right now. According to what I want (longevitity over OC), what speed do you suggest I run at?



So you're suggesting that the 60-66C temps while encoding video could possible be a problem with the temp sensors on the mobo?

If so, that makes it even harder to make sure where I should stay at the OC, or not to OC at all, even... What's your advice on this? Go back to 2.8Ghz, or try 3.0Ghz instead?

Mosfets are the little black capacitors you see around the CPU area. As opposed to the large tower types.

If you want longevity, clock down and see what your temps are like.

Temp sensors can be inaccurate. That's a fact. It's been seen many times. Touch the heatsink when it is under load and see if you can determine whether 66c would be accurate or not.

[Jon Lai]

If you want longevity, clock down and see what your temps are like.

Temp sensors can be inaccurate. That's a fact. It's been seen many times. Touch the heatsink when it is under load and see if you can determine whether 66c would be accurate or not.

I'm clocked down to 3.01Ghz now and it seems like idle temperatures are the same, rated at rarely 46C, mostly 47-49C while having non-CPU intensive programs running in the background.

However, it seems like the CPU temperature has yet to decrease. It still runs its way up to 70C - right now I'm 6.5 minutes into an approximately 30 minute video conversion, and the CPU temp has stayed at 68C for the last minute, when the peak for 3.22Ghz configuration was 70C.

edit: It eventually climbed to 70C around 7.5 minutes, but drops back pretty quickly.

edit2: I'm seeing random 71C's now around 8.5 minutes into conversion, and dropping back and forth... For my previous test clocking at 3.22Ghz, I stopped it at around 4 minutes.

edit3: 72C at 9.5 minutes into conversion, dropped back to 71C within 5-10 seconds (PC Probe is set to check numbers every 5sec)

If these temps are parallated, then it means that I must've ran the system to at least 66C when the computer was clocked at 2.8Ghz...

Mobo temp has been staying pretty much between 39C and 40C however. Perhaps I should go back to 2.8Ghz (stock clock)?

FYI, according to Task Manager in Windows XP, my video conversion process takes 95-98% of the CPU Usage... could that be taken into account?

[goofball]

If you haven't already, run the computer with the side panel off. See if that causes any decreases/increases in temperatures.

[Jon Lai]

If you haven't already, run the computer with the side panel off. See if that causes any decreases/increases in temperatures.

Nope, it seems like opening the side panel does not affect the CPU temp. I left it off for around 2 minutes.

While I was at it, I tried touching the heatsink and the chipset on the mobo. The Intel chipset is not hot at all, I could leave my finger there forever and not be burned, however, the heatsink was quite hot. With that, I persume that the temperature sensors are more or less correct than not.

Right now, conversion is complete, and temp is slowly decreasing from the 71-72C during conversion, 52C right now.

Sounds like I'll need to clock down to 2.8Ghz... but according to the temp stats I'm getting, could it be possible that the temps were pretty high to start with (at stock clock)? Video conversion is an important factor of why I use my computer, FYI, and so in 10 hours of PCing, probably 2-3 hours are dedicated to video conversion.

BTW, I forgot if I mentioned this or not, but I have only one case fan, it's an Antec and set to exhaust.

[HUH]

You can try set it back to default settings (fsb, voltages, timings, etc..) and see what temp you are getting after some video conversion. If it's still in the 60/70s, then probably your case airflow is not good enough, or your ambient temp is already high. Can you also list your PSU and case? Generic? Is this a new system? is it dusty?

As someone already mentioned, those sensor only gives you a rough estimate of the actual temp, so don't trust it 100%.

[Jon Lai]

You can try set it back to default settings (fsb, voltages, timings, etc..) and see what temp you are getting after some video conversion. If it's still in the 60/70s, then probably your case airflow is not good enough, or your ambient temp is already high. Can you also list your PSU and case? Generic? Is this a new system? is it dusty?

As someone already mentioned, those sensor only gives you a rough estimate of the actual temp, so don't trust it 100%.


I haven't evne touched the voltages and timings, only switched the FSB up to OC. Is that not a very good way to OC? Timings of my RAM went up automatically once I OCed, however.

My case is generic, yes, and PSU isn't the best either. It's brand is ORION 300W, but I'm sure 99% of us would consider that a "generic" PSU. I'm also running stock CPU fan.

I've always known that my cooling is probably just on the margin of being good. When I bought my system, mind you, it didn't even come with any case fans, only the CPU and PSU fans were running. It wasn't until I got my second hard drive that I added an exhaust fan onto my case.

BTW, it seems like I was wrong about 3.01Ghz and 3.22Ghz idleing at the same speed - still running at 3.01Ghz, I'm getting average 42C-44C idle temps with the usual BT client and MSN running in the background.

If in the end it appears that it's actually bad airflow that's causing the problem, and not OC, even at stock 2.8Ghz clock, could I continue using my computer as normal? I've probably encodded 200+ hours worth of video over the last 1.5 years and my computer has served me well to this day *touchwood*.

And you see, I don't know how much I should trust the readings, but as I touched the heatsink myself, it does seem pretty high, so likeliness is that it's more correct than incorrect.

[goofball]

That's a very good way to overclock, if it is stable. Voltages = increased heat, among other things. Heat ruins everything.

I would maybe stop overclocking, and try to work on getting better cooling in your computer. As I just mentioned, heat ruins everything, and even things running at stock will have decreased longevity due to excessive heat.

Orion PSU's tend to use very small heatsinks and run hotter than other PSU's with the same wattage but better heatsinks internally.

[Jon Lai]

That's a very good way to overclock, if it is stable. Voltages = increased heat, among other things. Heat ruins everything.

I would maybe stop overclocking, and try to work on getting better cooling in your computer. As I just mentioned, heat ruins everything, and even things running at stock will have decreased longevity due to excessive heat.

Perhaps. A friend of mine also reminded me to clean out the possible dust from the CPU fan, something I'll need to do when I turn off my computer tonight. He also mentioned that if it runs stable at the high temperatures, then it's okay, but personally I still don't feel confident doing that.

Mind you, I'm new to OCing and so I wanted to test it out to see how it goes, so I'm not prepared to spend money upgrading cooling methods on my computer just to do that. Perhaps after cleaning off the dust from the stock Intel fan I'll get to more reasonable temperatures. He also said that 3.2Ghz is quite a reasonable clock for 2.8 Northwoods, so it may just be my cooling.

Orion PSU's tend to use very small heatsinks and run hotter than other PSU's with the same wattage but better heatsinks internally.

Oh, and I thought it was just another no-name.

I don't quite understand you though, you said Orion PSUs have small heatsinks, but they have better heatsinks internally? So their small heatsinks run better than other PSUs with big heatsinks?

BTW, I don't know much about Orion (and another reason why I thought it was generic PSU), do they actually make good PSUs?

[goofball]

Orion is not bad but it depends. It should be an HEC made unit, which is a pretty decent unit but they do have varying levels of quality, Orion being on their lower end.

The heatsinks in an Orion PSU tend to be small. I was comparing it to a PSU that would have the same wattage but better larger heatsinks internally, and how the other PSU would be better overall.

I personally wouldn't buy one without having had time to open the unit up and subject it to proper testing. There is just something about a 565w PSU for $60 regular price, that just doesn't sit well with me. I'm an RFD'er and cheap as hell but I still am wary.

[Jon Lai]

Orion is not bad but it depends. It should be an HEC made unit, which is a pretty decent unit but they do have varying levels of quality, Orion being on their lower end.

The heatsinks in an Orion PSU tend to be small. I was comparing it to a PSU that would have the same wattage but better larger heatsinks internally, and how the other PSU would be better overall.

I personally wouldn't buy one without having had time to open the unit up and subject it to proper testing. There is just something about a 565w PSU for $60 regular price, that just doesn't sit well with me. I'm an RFD'er and cheap as hell but I still am wary.

I still don't understand your comparison about the heatsinks thing, but I don't think it matters anyways. The PSU came with the system when I bought it (back then, I was a hardware n00b, and right now, I still am, still trying to learn exactly how I can make everything work properly).

Back on topic. I'll still have to clean off the dust tomorrow from my CPU stock fan and see if it'll bring my temp a few degrees down on load so that at least it doesn't hit the 70C mark, then I'll be more confident in higher OCs.

Is there really anything else I could possibly do without investing a good $40+ on a CPU cooler?

And back to my original concern - you have to realize that I was using this computer for 1.5years already UNDER THESE EXTREME LOAD TEMPERATURES (that I didn't know about until now). Was it a mistake to do this to start with, or could I conclude that since I've been using it for so long at these temps anyways, perhaps it's okay?

[goofball]

I still don't understand your comparison about the heatsinks thing, but I don't think it matters anyways. The PSU came with the system when I bought it (back then, I was a hardware n00b, and right now, I still am, still trying to learn exactly how I can make everything work properly).

And back to my original concern - you have to realize that I was using this computer for 1.5years already UNDER THESE EXTREME LOAD TEMPERATURES (that I didn't know about until now). Was it a mistake to do this to start with, or could I conclude that since I've been using it for so long at these temps anyways, perhaps it's okay?

Let's compare PSU 1 and PSU 2. PSU 1 weighs 5lbs and has 3lbs of heatsink inside. PSU 2 weighs 3lbs and has 1lb of heatsink inside. Which one do you think will be better in terms of heat dissipation, efficiency, and power output? I think that should clear up what I was trying to say. If you don't understand that, I think you should invest some time at Silent PC Review under their PSU section, as well as other places like Xbit Labs.

Let me ask you this: just because you were unaware, does that mean that you couldn't have possibly caused any damage? As I've stated before, heat destroys everything. It doesn't need to fail at this very moment, to have any sort of damage that may not show up now but over time. Like the foundation of a house. It takes a long time but it will get damaged. That's just the way it is.

Now, I'm not saying you did cause any damage. What I am saying, is that the life of your hardware has probably been shortened by the temps at which you were running your hardware at. Will it matter to you? I can't say. You may not have any problems during the life you own your stuff. Then again, you might. Such, is the way life is.

Try to get your temps down a bit, however you decide to do so. Clean up your wiring, maybe install a case fan or 2, drill a side intake fan, whatever.

[Jon Lai]

Nevermind, it looks like it was dust that was causing trouble. I did some housecleaning this morning and my idle temps magically dropped beyond my expectations to 31C-33C at 3.0Ghz. Testing full load under this clock first, right now by the looks of things it looks like it's staying at around 49C-51C under full load at 3.0Ghz. I might just give 3.22ghz another try after some more testing.