CPU Temps

[Bugbait]

I was contemplating using this method but I did a bit of research around the hexus forums and it seemed that the most popular way of applying the paste was to spread a thin layer on the cpu. Do you think its worth trying the instructed way? More importantly do you think it could result in lower temps?

If you have the time and patience it can't hurt to try it out. I've used the spread method for many years and the few times I tried other ways didn't result in better contact and a few times it was worse, maybe I'm just crap at it. I get consistently "expected" results using the spread method so continue to do so.

[Bugbait]

Thats the perfect way to describe the thickness of how I spread it Also the arctic website says something about 'break in time' something to do with the paste breaking in (lol) so does that mean that the temps could drop still after the break in time has been reached?

It's more a settling time I think. I have noticed slight drops but I don't track the delta between ambient, etc. scientifically so it may not change at all. Doesn't seem to get worse though, unlike some pure silicon TIM's. Even then, the degradation usually takes longer than I keep most CPU's untouched. Don't expect greater than a 5C, or even 3C (probably the realistic upper end) drop once it settles though. It's small, not so ground breaking you'll be boasting from the rooftops.

[KalC]

If you have the time and patience it can't hurt to try it out. I've used the spread method for many years and the few times I tried other ways didn't result in better contact and a few times it was worse, maybe I'm just crap at it. I get consistently "expected" results using the spread method so continue to do so.

Yup I see what you mean, the only reason I chose this method was that yes it was popular and also the knowing of that the paste has been spread properly all over the cpu...saying all this though il give the other way a shot tomorrow, see how it goes.

[Taz]

I was contemplating using this method but I did a bit of research around the hexus forums and it seemed that the most popular way of applying the paste was to spread a thin layer on the cpu. Do you think its worth trying the instructed way? More importantly do you think it could result in lower temps?

Well, that's how I put my HSF onto my Q6600 and i'm happy with the temperatures. It wouldn't hurt to experiment but I just used the advice of Arctic Silver themselves who, I would imagine, have tested all sorts of spreading methods and advised the non-spreading method as the ideal method for their thermal paste.

[Bugbait]

Well, that's how I put my HSF onto my Q6600 and i'm happy with the temperatures. It wouldn't hurt to experiment but I just used the advice of Arctic Silver themselves who, I would imagine, have tested all sorts of spreading methods and advised the non-spreading method as the ideal method for their thermal paste.

I think there's a component of "most fail safe" or "most consistent" in their recommendations. Makes sense, most people can consistently squirt out a straight line. Ask them to spread a "paper thin" layer and I bet it varies greatly with lumps and rubbish all over, not to mention the danger of dripping it down the sides of the CPU.

[Thorsson]

Ambient temps will affect the Antec 900 very directly, but it's not that much hotter currently than room temp in a CH home. Almost certainly the whole AS5 thang is a red herring. Undoubtedly the CPU is not perfectly flat, but I don't think that can account for more than a few degrees. If it was particularly bad then there could be a big difference between the cores. Is that the case?

What voltage is the CPU running at, given that power rises exponentially with voltage? What about dust on the 900 fan intakes?

[KalC]

Ambient temps will affect the Antec 900 very directly, but it's not that much hotter currently than room temp in a CH home. Almost certainly the whole AS5 thang is a red herring. Undoubtedly the CPU is not perfectly flat, but I don't think that can account for more than a few degrees. If it was particularly bad then there could be a big difference between the cores. Is that the case?

What voltage is the CPU running at, given that power rises exponentially with voltage? What about dust on the 900 fan intakes?

The cores show identical temps to each other currently at 42 degrees reading by coretemp on both..and i havent overclocked anything, just everything running at standard. CPU-Z is currently reporting 1.296V Vcore. I would say theres a light film of dust on the front side of the case but nothing to affect the temps majorly

[Thorsson]

Don't have an E6300, but 1.296v at stock seems a tad high to me, probably explaining why your idle temp is a little high. If it's on Auto then the draw may well increase under load.

[TAKTAK]

i use the 'grain of rice' method and yes temps are lower after a few heat cycles (leave computer on for overnight and turn off in the morning then turn back on later in the day) this allows the thermal paste to cure

but with the weather were having at the minute you would expect higher temps (mine have skyrocketed from 37 idle to 47 idle, and my mobos @ 40c atm -mainly due to the fact that i have a less that satisfactory cooler atm)

[Cosworth]

What program are using to read the temperatures.

Taking in to account as the others have said taking into the recent ambient temperature increase I too was still very surprised too see that Core Temp was now showing an increase in temperature from 38 degrees to 60 at idle. This I thought could not be right, so I checked it again with Everest and this now shows it to be only 43 degrees. So one them is wrong and my money is on Core Temp.

[KalC]

Don't have an E6300, but 1.296v at stock seems a tad high to me, probably explaining why your idle temp is a little high. If it's on Auto then the draw may well increase under load.

Really? So 1.296V is high on a system that hasn't been overclocked? I left the setting on Auto and haven't raised it...i just assumed it would give it the minimal amount to run

i use the 'grain of rice' method and yes temps are lower after a few heat cycles (leave computer on for overnight and turn off in the morning then turn back on later in the day) this allows the thermal paste to cure

but with the weather were having at the minute you would expect higher temps (mine have skyrocketed from 37 idle to 47 idle, and my mobos @ 40c atm -mainly due to the fact that i have a less that satisfactory cooler atm)

'grain of rice' method as in you dont spread the paste all over? im just about to try this way now and see how much of a difference (good or bad) it makes

What program are using to read the temperatures.

Taking in to account as the others have said taking into the recent ambient temperature increase I too was still very surprised too see that Core Temp was now showing an increase in temperature from 38 degrees to 60 at idle. This I thought could not be right, so I checked it again with Everest and this now shows it to be only 43 degrees. So one them is wrong and my money is on Core Temp.

I too have checked with both everest and coretemp and they both show the same reading which is currently at 42 on both cores

[TAKTAK]

'grain of rice' method as in you dont spread the paste all over? im just about to try this way now and see how much of a difference (good or bad) it makes

yeh, it's just simply as the name suggests, a blob of TP the size of a grain of rice in the middle of the Heat Spreader, then plonk the heatsink carefully down so that it spreads evenly (i tried all of the methods, and this is the quickest and it seems to give the same results )

[moogle]

maybe the IHS on the cpu is not 100% flat?

what do you do if the IHS is not flat and it's causing temp problems. RMA?

[KalC]

yeh, it's just simply as the name suggests, a blob of TP the size of a grain of rice in the middle of the Heat Spreader, then plonk the heatsink carefully down so that it spreads evenly (i tried all of the methods, and this is the quickest and it seems to give the same results )

I just tried that method now and the temps are identical...the only thing I gained was cut fingers (damn ninja) but im hoping after its 'break in' time that I see lower temps..well here's hoping

[Thorsson]

With these chips every different CPU has a different stock voltage. Yours seems to be pretty high. Checking around others are fairly close to 1.2v while you're at nearly 1.3. That means you're producing ~15% more power, and that translates into heat. Obviously the current hot spell in the UK is adding another few degrees.

I'd be interested to know how it's affected Blitzen's temps. He has a 900 and a Scythe Mine.

Oh, and I've never seen bedding in make more than a degrees difference. Hope you do better.

[KalC]

With these chips every different CPU has a different stock voltage. Yours seems to be pretty high. Checking around others are fairly close to 1.2v while you're at nearly 1.3. That means you're producing ~15% more power, and that translates into heat. Obviously the current hot spell in the UK is adding another few degrees.

I'd be interetsed to know how it's affected Blitzen's temps. He has a 900 and a Scythe Mine.

would there be any way to lower the voltage even though its on an auto setting? presumably through bios?