Choosing a CPU cooler

[Nec_V20]

So you have been looking around reading reviews, watching YouTube and you are pretty confused with regard to CPU coolers.

What's all the fuss about? Simply put, the hotter the CPU runs the shorter its lifespan will be. One should not be led astray by anyone talking about "safe" temperatures. Your CPU will die at some point as the heat generated by it burns it up, it is the coolers job to drag out that inevitability. There is also no "safe overclocking". A CPU with a temperature of 48 degrees Celsius will last longer than one running at 60 degrees. Anyone telling you otherwise is just - how do I put this politely - lying. There is an exception (see "Silicon Lottery" below).

There is something known as "The Silicon Lottery" which means that your CPU will be randomly more or less resistant to failure (heat death). If you do overclock then one CPU may run stably at a certain speed whereas another will just crash. There is no way of telling short of actually using the CPU if you are a winner or a loser in this lottery. You also cannot send a CPU back the manufacturer which will not run stably at even a mild overclock because technically you are running it outside the manufacturers specification.

If you have an Intel CPU then the cheapest way to knock a few degrees Celsius off the CPU temperature is to turn off HyperThreading in the BIOS.

There are two kinds of coolers and those can be split into two general categories:

1) Air Coolers
a) Stock Coolers
b) Third Party Air Coolers

2) Water Coolers
a) Custom Parts Water Coolers, Liquid Nitrogen
b) AIO (All In One) Water Coolers.

The easiest ones to discuss are those falling into the category 1a, the Stock Coolers. These are the ones you will get with your CPU when you buy it. The only thing that can be said about them is, "Better than nothing, but not by a lot". In other words they are garbage and should be consigned to the bin at the earliest possible opportunity. From my own recent experience, I bought an AMD A8-5600k processor for my NAS and I put the stock cooler that came with it onto it to see how it would perform before putting a real cooler on it.

I booted up the system, started the hardware monitor to check the CPU temperature then I started Prime95 to put the CPU under load. Within seconds after starting Prime95 the CPU temperature shot up to 62 degrees Celsius, which is totally unacceptable and I stopped the testing immediately so as not to damage the CPU.

The next easiest to discuss would be the ones fitting into the category 2a or "Custom Parts Water Coolers". For this solution you have to buy all the pumps, radiators, reservoirs, tubes, joints etc. yourself and put it all together. These will give you the greatest cooling of your CPU if you have done it right; however they do have major drawbacks:

1) You will have to permanently monitor the system because they are prone to failure and leaks.
2) You will have to run all kinds of routine maintenance on them from checking the fill level of the coolant to making sure that all the joints are tight and not leaking, check for algae.
3) It is exorbitantly expensive
4) You will have to sacrifice a significant portion of the expandability of your chassis to the components of the cooling subsystem (coolant reservoirs)
5) They are only as efficacious as the builders ability to construct the complete system.
6) No warranty and dubious reliability. This solution has a VERY limited field of application (especially Liquid Nitrogen which has the added downside of being extremely dangerous if the LN is handled incorrectly).

The next two, 1b "Third Party Air Coolers" and 2b "AIO (All In One) Water Coolers" are the ones that the overwhelming majority of people will choose from. This is also the portion of the post where I will get the most grief.

Let's start off with Third Party Air Coolers. These range from only slightly better than the stock coolers but with advantages like having a smaller dimension. The one thing which can generally be said is that the lighter (in weight) the cooler is, the worse the cooling performance will be. These types of coolers are also, if one is not looking at overclocking, relatively cheap and one can buy one which will fit any kind of chassis one might choose to put the system into. The cooling performance ranges from good enough to great. The "great" will depend on the size of your chassis and your wallet.

How well an air cooler performs is heavily dependant on things other than the design or efficiency of the cooler itself. Essentially what the air cooler does is that it takes heat away from your CPU and dumps it into your case. It is up to the case ventilation to get rid of that heat. In a poorly ventilated case essentially what happens is that warmer and warmer air is recycled and used to "cool" the fins of the air cooler. The other thing is that a higher ambient temperature within the case will also effect the cooling of other components on the motherboard or the graphics adapter.

Then there is the elephant in the room - literally. To achieve the best cooling of the CPU air coolers have to be massive. All their weight is supported by the motherboard and generally they are oriented horizontally to the vertical alignment of the motherboard. This means that if you are moving your system and you drop it then there is a chance that the cooler will break the motherboard. This is apart from the permanent strain the weight of the cooler puts on the motherboard.

Then there is the design of the cooler itself, or rather the mounting system the manufacturers choose. These range from very easy and well thought out to the feeling that the manufacturers of the cooler are bloody-minded sadists hell bent on making your life a misery. Then there are the coolers which have a clip on system for the fans, however if the clips ever spring open for some reason there is a good chance they will short-circuit your graphics card.

Another thing to take into account is whether or not the components UNDER the CPU cooler will get any ventilation or that it effectively insulates the capacitors, VRMs around the CPU from any airflow which might cool them. If a capacitor blows due to excessive heating then it is game over for your motherboard.

If you are going to choose an air cooler for the CPU then you will have to plan the choice of things like RAM with regard to that choice. In many cases RAM slots will be blocked to anything other than low profile RAM. So you could end up having bought the cooler, bought your RAM and end up not being able to put your system together. Another scenario could be that you have bought the components and although the RAM fits, you won't be able to put the side panel onto the chassis because of the height of the cooler.

One advantage of the air coolers is that even if the fan does give out there will still be some cooling of the CPU. OK that is a fat lot of good if you are gaming or have overclocked the system up the wazoo; however if you are just doing normal stuff like browsing the Internet and posting to this forum or you are writing a paper with your office package you can get away with not having to immediately replace the fan.

Personally I would say that if you are considering an air cooler then the gold standard against which you should judge the competition are the Noctua series of coolers. As I have tried to point out, air cooler benchmarks with regard to how well they cool the CPU are not the entire story - by a very long shot.

Last but not least we come to 2b "AIO (All In One) Water Coolers". These kind of coolers have a small footprint on the CPU itself and remove the heat via the liquid to a radiator to be disposed of as you see fit. The main advantage of the water cooler in my estimation is that the weight of the business end is taken up by a component of the system which is specifically designed to take the stress of the weight, namely the chassis itself - if you drop your system then you might dent your chassis, but you are unlikely to destroy your motherboard. No air cooler other than the highest end ones even come close to giving the cooling performance of an AIO water cooler. The highest end air coolers are also as or more expensive than a water cooler offering the same performance.

The one thing one thinks about with regard to water coolers is "leaks". This possibility has been pretty much eliminated with AIO water coolers and I have not heard of any such failure - although I am sure such a thing has occurred.

So the choice is a no-brainer right?

Wrong.

Water coolers are expensive and if you only need something which is good enough or have a processor and system which is energy efficient then it is hard to justify the expense of a water cooler. Even though mounting the radiator to the chassis is the best place for the business end of any cooler to be there are many systems where there is just not enough room to do so. If you want to build a system as small as possible then this will be a deal breaker with regard to a water cooler.

If the pump on the water cooler fails then you will not be able to repair it yourself. With an air cooler, if a fan fails then it is very simple and relatively cheap to replace it and restore the cooler to its former glory.

The other problem one has with regard to water coolers is the tubes taking the liquid from and to the pump. Even if you do have a spot to attach the radiator to the chassis you might not be able to get the tubes to cooperate. Jamming the water cooler into your chassis and placing stress on the joints where the tubes connect to the radiator or the pump is a recipe for disaster if even one of those joints ever give and a leak occurs. If you kink the tubes on the water cooler then the coolant will not be able to circulate and your temperatures will shoot up.

I have tried to be as fair as I can with regard to the choices one has with regard to cooling a CPU and I hope this doesn't disintegrate into a series of fanboy rants.

 

[lehtv]

A word of advice: no one is going to read through all that. I suggest you edit the post and summarize your question to a few paragraphs (if it even is a question, I'm not sure).

 

[Nec_V20]

A word of advice: no one is going to read through all that. I suggest you edit the post and summarize your question to a few paragraphs (if it even is a question, I'm not sure).

Yes, let's shorten it and then have everyone complain about what I did not put into the post.

That's a no-win proposition.

After I had originally written the post I did in fact cut out a lot and reduced it to what I considered to be the bare essentials to give someone a fair general guide as to what is out there.

By being fair I mean that if you read the post I do not think you can discern which kind of cooling solution I have chosen for my systems (except of course to dump the stock cooler).

 

[dma0991]

What's all the fuss about? Simply put, the hotter the CPU runs the shorter its lifespan will be.

Hotter does mean shorter lifespan but if it is within the manufacturer's specifications, it will last the usable lifetime of the PC.

Anyone telling you otherwise is just - how do I put this politely - lying. There is an exception (see "Silicon Lottery" below).

Silicon lottery exists but not to the average user who never runs their processors out of the manufacturer's specifications. Even with K series processors, Intel is not responsible for guaranteeing that your CPU will reach 5GHz as they only promised according to the advertised specs.

I booted up the system, started the hardware monitor to check the CPU temperature then I started Prime95 to put the CPU under load. Within seconds after starting Prime95 the CPU temperature shot up to 62 degrees Celsius, which is totally unacceptable and I stopped the testing immediately so as not to damage the CPU..

TJmax for an AMD A8-5600K is 74C, so 62C is in no way damaging to the CPU as it is within specifications. Another thing is that since Llano, AMD's APUs have always had higher max voltage than it really needs. There have been cases where undervolting is done to the stock APU which reduced the load temperatures by a lot and does not compromise its stability.

The next easiest to discuss would be the ones fitting into the category 2a or "Custom Parts Water Coolers". For this solution you have to buy all the pumps, radiators, reservoirs, tubes, joints etc. yourself and put it all together. These will give you the greatest cooling of your CPU if you have done it right; however they do have major drawbacks:

With good assembly procedures and patience, a lot of leaking issues with custom water cooling can be easily mitigated. Its expensive, there's no doubt about it, except that it is classified as more of a hobby for enthusiasts. There are no cheap hobbies for an adult, except if stamp collecting is your thing.

This means that if you are moving your system and you drop it then there is a chance that the cooler will break the motherboard. This is apart from the permanent strain the weight of the cooler puts on the motherboard.

You underestimate the strength of PCBs. I have scrap PCBs used in the production of PSUs and they have a great amount of flex before snapping. I have never seen a PCB break due to a heavy heatsink but I've seen it warp, caused when the CPU cutout wasn't big enough and the back plate is resting on the motherboard tray, causing it to bow.

If you want to build a system as small as possible then this will be a deal breaker with regard to a water cooler.

On the contrary, watercooling is great for small systems. Take this as example of a well executed SFF watercooling. Temps are just as impressive, as it cools both CPU and GPU with a single thick 120mm radiator. SFF watercooled builds can be done with AIOs, particularly ones with thin 120mm radiator.

If you kink the tubes on the water cooler then the coolant will not be able to circulate and your temperatures will shoot up.

One of the selling points since AIO first appeared is that how the tubing almost never kinks. If it kinks, its PEBKAC.

 

[lehtv]

Yes, let's shorten it and then have everyone complain about what I did not put into the post.

That's a no-win proposition.

After I had originally written the post I did in fact cut out a lot and reduced it to what I considered to be the bare essentials to give someone a fair general guide as to what is out there.

By being fair I mean that if you read the post I do not think you can discern which kind of cooling solution I have chosen for my systems (except of course to dump the stock cooler).

It's a guide? I thought you are choosing a CPU cooler and asking for input. Maybe you should rename the title to avoid confusion, e.g. "Guide to choosing a CPU cooler".

Even so, my opinion is that a guide to choosing a CPU cooler can be perfectly well laid out in a few paragraphs, without need for further explanation. But even if you disagree about that, I suggest making it more readable. Don't take me wrong, I'm not trying to whine or complain, I think it's great that people write guides. But you could improve the transfer of your ideas by headings and subheadings, and providing a summary in the beginning. The post is just so long that it comes off as a wall of text otherwise.

 

[Nec_V20]

One of the selling points since AIO first appeared is that how the tubing almost never kinks. If it kinks, its PEBKAC.

Never underestimate the power of PEBKAC or the ID-TEN-T situations one is confronted with even when doing tech-support at Enterprise level where you would think you are dealing with people who know what they are doing.

In fact I am just writing a post now which documents my own PEBKAC (I prefer writing it PEBCAK) and how I managed to ah heck up my motherboard within a half an hour of turning on the new build. I did manage to get it sorted though.

On the contrary, watercooling is great for small systems. Take this as example of a well executed SFF watercooling. Temps are just as impressive, as it cools both CPU and GPU with a single thick 120mm radiator. SFF watercooled builds can be done with AIOs, particularly ones with thin 120mm radiator.

The one you documented is indeed very nice, however I was trying to keep the post as generic as possible and if I had included all the "exceptions prove the rule" examples that I could think of then the post would have been even longer than what leytv was complaining about.

I did cut out quite a lot from the original draft of the post and obviously those omissions are a point of criticism. However you have done a good job of adding meat to my original post.

Although your point is well taken, I think that the temps on my A8-5600k going to 62 degrees Celsius within seconds of starting Prime95 on the system is a good reason to dump the stock cooler.

And when I say within seconds it was, click on prime95 to start, then click down at the bottom of the screen to bring up the monitoring software and seeing that the temperature was at 62 degrees.

Hotter does mean shorter lifespan but if it is within the manufacturer's specifications, it will last the usable lifetime of the PC.

To be fair to you, I have never had a CPU die on me before I decided that I had to upgrade to a higher performance system anyway. I have had motherboards die on me before the effective end of life of the system was reached. The most expensive case of this was when my Tyan Tomcat board died on the dual Pentium MMX 200 system I built and had to replace it.

All in all though your points were fair and the criticism is well taken - even if I do disagree on some things.

 

[Nec_V20]

It's a guide? I thought you are choosing a CPU cooler and asking for input. Maybe you should rename the title to avoid confusion, e.g. "Guide to choosing a CPU cooler".

Even so, my opinion is that a guide to choosing a CPU cooler can be perfectly well laid out in a few paragraphs, without need for further explanation. But even if you disagree about that, I suggest making it more readable. Don't take me wrong, I'm not trying to whine or complain.

Do I get brownie points for the fact that English is my second language (German is my native language)? :'(

I think it's great that people write guides. But you could improve the transfer of your ideas by headings and subheadings, and providing a summary in the beginning. The post is just so long that it comes off as a wall of text otherwise.

Personally I don't do "Death by Powerpoint". With regard to "summary in the beginning" it would not only have made the post longer than what you complained about but also people would have been rushing to criticise the summary for the parts I did not include which were actually in the body of the text they did not bother to read.

I think that the original post was concise enough to be readable and yet long enough to be useful. Your criticism focusses in on form over substance and we will just have to agree to disagree on that point because I write the way I do and you have your own style which you prefer.

 

[lehtv]

Do I get brownie points for the fact that English is my second language (German is my native language)? :'(

Uh no. Your English is fine. I'm not talking about grammar or sentence structure, although it's probably true that it is easier to write concise yet precise text in your own language. Nevertheless my point was about it looking like a big wall of text instead of a more easily approachable piece with sections or subheadings.

Personally I don't do "Death by Powerpoint". With regard to "summary in the beginning" it would not only have made the post longer than what you complained about but also people would have been rushing to criticise the summary for the parts I did not include which were actually in the body of the text they did not bother to read.

PowerPoint? I never suggested using bullet points...

I said I wasn't complaining... And I'm still not complaining. So don't say I'm complaining OK?

The point of a summary is that it's a separate text from the rest. It does technically make your post longer, but it also makes it more interesting. People who read the summary and find it interesting are much more likely to actually continue reading to find out the details, compared to people who don't have a summary to read in the first place. Also, a summary is not a text that contains some parts and doesn't contain other parts of the full text. It outlines what the text it about without giving any details, and as such it doesn't leave anything important out.

Of course, it's a matter of taste when a text is long enough that a summary is useful.

I think that the original post was concise enough to be readable and yet long enough to be useful. Your criticism focusses in on form over substance and we will just have to agree to disagree on that point because I write the way I do and you have your own style which you prefer.

I am not talking about what style I write with, I'm talking about what style I would actually read, and what I think others would also read. And not just read, but get as much out of as possible.

 

[T_Yamamoto]

To be honest. No one is going to read what you wrote. The average Joe isn't going to care about this post. This post will eventually disappear and no one will notice. (Harsh, but I'm being honest)

 

[Nec_V20]

To be honest. No one is going to read what you wrote. The average Joe isn't going to care about this post. This post will eventually disappear and no one will notice. (Harsh, but I'm being honest)

You don't care about my post and have no use for it, so therefore you speak for all others who may ever read my post? And in which universe do you live that you are the arbiter for all other people of all things worth caring about?

When I write something like my original post I do so remembering how lost I felt when I first started off with computers. There was no Internet, there was hardly any published material that I could readily get hold of to resolve problems that I had. I did however have the incredible luck of two or three techies taking me under their wing and helping me.

To this day I remember how happy I was to have something explained to me and the help that gave me. This is what I had in mind when I wrote the post. It is payback - maybe 30 years later - for the effort those people invested in me.

There are a lot of people who think to themselves "What the eff is all this cooling about and why should I bloody well care?" I tried to approach that question from a non-fanboy standpoint.

Why the hell should you care whether anyone reads what I wrote or not? Nobody forced you to read it, and if you did then there must have been something in the way I wrote it that kept your interest long enough to actually comment - no matter how shallow that comment was. So from that standpoint alone my work here is done.

However on the off chance that it helps even just one other person at some point now or in the future make some sense of things then I have paid back a debt that I owe to those people who helped me 30 years ago.

 

[Nec_V20]

Uh no. Your English is fine. I'm not talking about grammar or sentence structure, although it's probably true that it is easier to write concise yet precise text in your own language. Nevertheless my point was about it looking like a big wall of text instead of a more easily approachable piece with sections or subheadings.

PowerPoint? I never suggested using bullet points...

I said I wasn't complaining... And I'm still not complaining. So don't say I'm complaining OK?

The point of a summary is that it's a separate text from the rest. It does technically make your post longer, but it also makes it more interesting. People who read the summary and find it interesting are much more likely to actually continue reading to find out the details, compared to people who don't have a summary to read in the first place. Also, a summary is not a text that contains some parts and doesn't contain other parts of the full text. It outlines what the text it about without giving any details, and as such it doesn't leave anything important out.

Of course, it's a matter of taste when a text is long enough that a summary is useful.

I am not talking about what style I write with, I'm talking about what style I would actually read, and what I think others would also read. And not just read, but get as much out of as possible.

Did I get anything substantively wrong in your opinion? is the forum here worse off for my having posted what I did?

 

[Agent11]

Why would you turn off hyperthreading? You realize you pay extra for a cpu with this feature right?

 

[T_Yamamoto]

You don't care about my post and have no use for it, so therefore you speak for all others who may ever read my post? And in which universe do you live that you are the arbiter for all other people of all things worth caring about?

When I write something like my original post I do so remembering how lost I felt when I first started off with computers. There was no Internet, there was hardly any published material that I could readily get hold of to resolve problems that I had. I did however have the incredible luck of two or three techies taking me under their wing and helping me.

To this day I remember how happy I was to have something explained to me and the help that gave me. This is what I had in mind when I wrote the post. It is payback - maybe 30 years later - for the effort those people invested in me.

There are a lot of people who think to themselves "What the eff is all this cooling about and why should I bloody well care?" I tried to approach that question from a non-fanboy standpoint.

Why the hell should you care whether anyone reads what I wrote or not? Nobody forced you to read it, and if you did then there must have been something in the way I wrote it that kept your interest long enough to actually comment - no matter how shallow that comment was. So from that standpoint alone my work here is done.

However on the off chance that it helps even just one other person at some point now or in the future make some sense of things then I have paid back a debt that I owe to those people who helped me 30 years ago.

95% of the people won't care.

Better?

Wow, you're really mad.

 

[Nec_V20]

Why would you turn off hyperthreading? You realize you pay extra for a cpu with this feature right?

If you are overclocking then hyperthreading adds about four degrees Celsius just at normal usage or what one could call "idle". That penalty in heat rises substantially as you put your system under load and increase the clock and the voltages above stock.

In gaming, most games cannot effectively use multiple cores; and by doubling the amount of cores through virtualization you are exacerbating the amount that your Cache gets dirtied and the CPU having to load data from the very much slower RAM, thus losing performance (you get the penalty of the Cache miss plus the penalty of having to go to RAM).

Hyperthreading does add a lot of value, but also subtracts from it. It is no panacea for every situation.

And could you please be even more condescending in your next reply?

 

[Nec_V20]

95% of the people won't care.

Better?

Wow, you're really mad.

Why don't you just be honest and say that you don't care?

And would that be 95.49% of people or just 94.51%.

Why should it ever be any more than just yourself that you are speaking for?

You didn't like it, so OK - it's not like I could, or would, go back and delete it.

 

[Nec_V20]

So far only dma0991 has added something constructive to this thread which I started with my post.

The only thing that the others need is some cheese with their whine.

 

[TerryMathews]

Point of order, not all stock heatsinks are garbage. The FX-8350 retail heatsink, for example was a really nice Cooler Master heat pipe unit.

 

[Nec_V20]

Point of order, not all stock heatsinks are garbage. The FX-8350 retail heatsink, for example was a really nice Cooler Master heat pipe unit.

And I would echo "point of order" to you sir. How many other CPUs does AMD package in a metal box?

Let's face it, it needed something special in the way of "stock" cooling because it was way overclocked and would have fried if AMD had delivered it in the same way they do any other of their CPUs.

Even with that twin heatpipe pathetic little "cooler" it would have gone tits up as soon as Prime95 started.

Under load it shoots up to 65 degrees Celsius from an idle temp of 30 degrees Celsius. How long do you think you could have kept Prime95 running before the processor disintegrated even with the "really nice" stock cooler?

Under load it consumes 227 Watt! A comparable Intel CPU with higher specs on benchmarks, the Core i5-3570k, only consumes 122 Watt.

A lot of that extra energy being pumped into the FX-8350 to achieve the bragging rights of 4.0 GHz (4.2 Turbo) is bought at the expense of a heck of a lot of heat when all eight cores are put under load.

As I said to dma0991, "... if I had included all the "exceptions prove the rule" examples that I could think of then the post would have been even longer ...".

The cooler supplied with the FX-8350 would have looked OK on an A10-6800k Richland but on the FX-8350 it is like spitting on a hotplate.

 

[TerryMathews]

And I would echo "point of order" to you sir. How many other CPUs does AMD package in a metal box?

Let's face it, it needed something special in the way of "stock" cooling because it was way overclocked and would have fried if AMD had delivered it in the same way they do any other of their CPUs.

Even with that twin heatpipe pathetic little "cooler" it would have gone tits up as soon as Prime95 started.

Under load it shoots up to 65 degrees Celsius from an idle temp of 30 degrees Celsius. How long do you think you could have kept Prime95 running before the processor disintegrated even with the "really nice" stock cooler?

Under load it consumes 227 Watt! A comparable Intel CPU with higher specs on benchmarks, the Core i5-3570k, only consumes 122 Watt.

A lot of that extra energy being pumped into the FX-8350 to achieve the bragging rights of 4.0 GHz (4.2 Turbo) is bought at the expense of a heck of a lot of heat when all eight cores are put under load.

As I said to dma0991, "... if I had included all the "exceptions prove the rule" examples that I could think of then the post would have been even longer ...".

The cooler supplied with the FX-8350 would have looked good on an A10-6800k Richland but on the FX-8350 it is like a piece of spit on a hotplate.

Have you ever used an 8350?

 

[Nec_V20]

Have you ever used an 8350?

I have never used a vast majority of the CPUs that have ever been produced.

Do you have another non sequitur you would like to pass by me?

 

[kiriki]

You can make this much easier to read by dividing the post up into sections like here and here.

 

[Nec_V20]

You can make this much easier to read by dividing the post up into sections like here and here.

I could have made it a damned sight easier by not posting it in the first place

Do you have any substantive dispute with what I wrote or do you just want me to send you a piece of cheese?

 

[dma0991]

Never underestimate the power of PEBKAC or the ID-TEN-T situations one is confronted with even when doing tech-support at Enterprise level where you would think you are dealing with people who know what they are doing.

I do understand that sometimes there are some who does things beyond comprehension or logic, and you can only ask "How did this even happen?". I knew such situation would happen when I built my friend's SB rig. He was adamant that he wanted an AIO but I insisted on a CM Hyper212+ instead because I know for a fact that maintenance is not a word in his dictionary.

Although your point is well taken, I think that the temps on my A8-5600k going to 62 degrees Celsius within seconds of starting Prime95 on the system is a good reason to dump the stock cooler.

Prime95 is not an accurate litmus test to know the upper limit of its operating temperatures. When I run my Core i5 3570K @ 4.5GHz, using Prime95 I get about 75C on average. Using a real application such as a F@H client, I get about 65C or less. Both program pegs the CPU load to a constant 100% but different temp results. I would prefer the reading of a F@H client as it does meaningful work, unlike finding prime numbers for the sake of stability testing.

If you are overclocking then hyperthreading adds about four degrees Celsius just at normal usage or what one could call "idle".

4C isn't a lot when HT does add to the final performance benefit. My Core i7 2600 runs about 55-60C at 100% load on a stock cooler, still within its safe limits.

 

[Nec_V20]

I do understand that sometimes there are some who does things beyond comprehension or logic, and you can only ask "How did this even happen?". I knew such situation would happen when I built my friend's SB rig. He was adamant that he wanted an AIO but I insisted on a CM Hyper212+ instead because I know for a fact that maintenance is not a word in his dictionary.

I have you beaten hands down on that one. I have a friend who shall remain anonymous - Hi Jim - who once phoned me up to tell me his computer was freezing.

Now when I originally installed the machine I had told him to stay away from dodgy websites and on the phone he swore blind to me that he had not been on any. I got there and the first thing I saw was that his desktop was filled with shortcuts to porn diallers.

I don't mind someone lying to me, but not THIS crassly.

He had originally gotten a message that there was a virus on his system but it could not be removed by the AV software that was installed. So he had downloaded another antivirus software and ran that and that also gave him the message that the virus could not be removed.

To make a long story short he had installed a total of five other AV software packages and after less than a minute of the system running after boot it froze up. The reason for this, I found, was that all six of the AV programs were doing real time scans at the same time. He wouldn't allow me to wipe his system because he had really important data on it and he wasn't sure if he would get it all if I just booted up with my USB flash drive and copied the stuff over. So I spent over six hours or more, booting up, uninstalling the AV packages he had put on there. He then spent the next couple of days making sure he had copied all his data over to the USB flash drive and then I wiped his system and reinstalled it again.

I'm sure that you and most others reading this must have a "Jim" in their lives who make one wonder why homicide is a criminal offence; I mean one is allowed to put horses out of their misery after all.

I like Jim and he is an intelligent person (he has a degree in Law) but as far as computers are concerned he has been a bane of my existence.

 

[TerryMathews]

I have never used a vast majority of the CPUs that have ever been produced.

Do you have another non sequitur you would like to pass by me?

I was just wondering.

I have, in fact, used the hardware we were talking about. And I can assure you it can Prime95 indefinitely. There is no thermal limit that you were implying.