How it's made: heat sinks

Because you're wrong, if you have two 6 inch fans on your water cooler radiator, you have the same air to heatsink cooling.

If they're not doing it to lower the temperature, they're not right in the head.

Because he puts GPUs bang up against each other with no air gap, and wants the heat outside the house.

Saying that after you've responded to every point makes you a silly little child who has to get the last word in. Were you bullied at school?

Clearly there is no reason to think you are going to understand, but the fans are not the only factor in the equation of removing heat. If they were, they would just put the fans to blow on the CPU and a heat sink would not be needed. The design of the heat sink and fins are important to optimize heat removal from the CPU and transfer it to the air. That's what the water cooler or heat pipe do better than a simple, metal heat sink.

You are the one who just said the CPU runs at 90°C, so there's no reason to cool further.

As I expected. No point at all. You may be unwilling to learn anything, but I'm starting to think you actually are unable. I don't get why you can't understand that most people who buy water cooling and heat pipes for the hot rod CPUs is to cool them as much as possible, so they can overclock them.

Cheers

Phil Hobbs

If only computers were still desktops and not towers.... Who thought it would be a good idea to have stuff hanging off things?

Use an IR thermometer to test one of your own.

Best watercooler has no fans on the radiator, or the radiator is in another room.

The one on the right is s*it, it lets my i5-8600K reach 100C and thermally throttle if I use all 6 cores and the onboard GPU to the max.

Sometimes you just have no f****ng clue! Well, for a wide range of "sometimes".

If you're going to disagree with me, to prevent yourself from looking like a wanker, explain why you believe I'm wrong. I have such a thermometer and can very quickly tell the temperature of every part of a heatsink. Having 96 CPU cores and 8 GPUs, I tend to analyze temperatures a lot.

And the heatsink on watercoolers is no better than the ones directly on the CPU. The bottleneck is moving heat from fins to air. Traditional watercooling doesn't change this.

I've never seen a CPU get that high with decent air cooling.

That's a yes then. Do you have specs?

Then they are fools.

The fool here is Comander Kinsey. He sees the CPU at one end of the heat dissipating arrangement and the finned heat sink at the other dissipating the heat generated, and hasn't noticed that there's a thermal gradient from the CPU to the surface of the air-cooled heat sink.

The copper block next to the CPU get almost as hot as the CPU. A heat pipe soldered onto that copper block can carry a lot of heat (as the latent heat of fusion of vaporised water) with very little temperature drop and dump it on to a much larger condensing area at the other end of the heat pipe where it can get spread across a fairly wide area under a large number of air-cooled fins, none of which have to get all that hot to transfer their little bit of heat to the passing air.

More dissipating fin area means the CPU doesn't need to get as hot to get rid of the heat it is dissipating. This might have dumbed it all down enough to make it comprehensible to Commander Kinsey, but since he isn't all that interested in understanding what's going, comprehension isn't his favourite activity. Poncing around telling people that stuff doesn't really matter is more his line.

It's not a bottleneck. It's a resistance. Thermal flow is exactly like resistance in an electrical circuit. The total resistance is the sum of the individual series resistances. You have no basis for claiming the thermal resistance at the metal to air interface is significantly larger than the resistance elsewhere. So reducing the resistance of the heat flow to the fins already reduces the total thermal resistance. In addition, you are assuming the fins, fans and everything else are the same. There's no reason to think that. Moving the bulk of the heat sink off the CPU means it can be designed without restriction to the weight. Noting the fans have the same diameter blades does not make the whole thing equivalent.

I know you aren't going to understand this and are going to be in denial about it. But, whatever. You clearly have your intellectual limitations.

Now you are doubling down on the idea. ;

LOL

Yes, anyone who does things you don't, is a fool. That cuts both ways.

Probably removed by the health and softy committee in case you reverse at high speed in to a pedestrian.

Donald Trump's car drifting in reverse very fast:

I know very well what it is, I have a f****ng physics degree. But bottleneck is a reasonable colloquial term for anything which is the main cause of a reduction in something. Think of traffic flowing along a highway then through a one lane roadworks, that's a bottleneck. It's also a resistance.

Basic physics tells us it is. Air doesn't conduct like solid.

You're adding more resistance, metal to water than back to metal again.

Anybody who uses the term "in denial" is probably one those greenies who thinks logical people are denying their hippy warming movement.

WTF? I'm pointing out air cooling keeps it well below the max.

That's also a yes. Electronic analysis of this thread determines this has a 76% chance of being you:

No it doesn't. You can't just put in cliches and make them apply to me.

There's a reason your 4GHz CPU isn't sold at 4.4GHz. Because it's not reliable at that speed.

It's hard to find a cliche that doesn't apply to Commander Kinsey. He is a very stereotyped wanker.

Actually its more that only some parts keep on working when clocked that fast. The specifications that manufacturers slap on a part are set by the production yield that they get when they test to that specification. They seem to want 95% yield or better, if they can get it, so the specifications tend to be conservative.

Over-clockers are interested in seeing how fast they can run the part they happen to have bought, with the chip cooler they decided to put on it.

When we put in select on test parts to compensate for production tolerances, parts from the same batch were a lot more similar to one another than parts from different batches (and we bought in batches that covered about six months production, so the guys doing the testing got pretty bored with putting in the same resistor for months on end). Parts that came from the same wafer might be even more similar.

If you have a physics degree, you would understand the differences between traffic and electrical resistance. So clearly I have caught you in a lie.

I don't need a poor analogy, because I understand resistance. Why not just work with thermal resistance, rather than talking in analogies that you clearly don't understand.

Lol! So that's physicist talk, "Air doesn't conduct like a solid"? The air in this situation doesn't need to conduct heat. It accepts heat from contact with the heat sink fins, and the movement of the air transports that heat away. Conduction has little to do with it other than possibly the very thin layer at the air-metal surface where it does conduct just like any other material.

Again, you can't understand that the remainder of the system does not need to be the same. You can't look at one portion of the system and talk about the entire system. You want desperately for the two systems to be otherwise identical, but there is no reason to believe that.

Or they are discussing something with someone who is in denial.

Which is an irrelevant point. People buy water cooling to get as low a CPU temperature as possible, so they can overclock as high a speed as they can get. It's like any other pursuit where the the goal is to optimize something, far beyond what is useful. It's a hobby.

You are right about the eyesight. Not so right about the muscles, etc. I was captain of the wrestling team and have been very successful in life. You, meanwhile, are reduced to the pathetic presence you project in this group, completely unable to have a meaningful discussion, largely devoid of friends.

Again, you fail to understand anything at all about this thread. You are an interesting case, but I think you have been explored enough. This is pretty much a hobby, and you are getting boring.

We aren't replying to you. We are jeering at you.

Sorry, it's too complicated for you. Imagine lots of cars going through a small road.

And what lie is this?

To make it simpler for you.

Yip.

By conduction. what do you think happens at the interface?

Stop waffling and get to the point.

No such thing, it's just believers that can't handle people who don't. Mainly religion, or some other "religion" like climate change.

If it could run faster, it would be sold as such. Do they really think the manufacturer doesn't know what they're doing?

A muscular geek, yeah right.

I see you suck at psychology too.

And yet again you supply no information, you just grumble when things aren't going your way. Are you a bot?

Can't argue with you there. They guy is like a few others here, no interest at all in learning anything as he already knows everything.

He literally doesn't understand the concept, plus, he's always afraid to admit he might be wrong and have something to learn. It would be such a blow to his world model with him at the center.

Whatever. There's no point in responding to a number of posters here. I wish GG had a kill file.