New Apple Pro.

[Pyromanik]

Thunderbolt for all the things Pyro.

It'd be nice if that were the case.

But "All the things" in this case is about 3.

That fan blows the wrong way too, all it's going to do is suck up all the dust off the floor...

yes and no, heat rises and all that.
Lots of modern cases put the PSU fan down there (sucking up, coz exhaust is out the back), but no one seems to complain about that.




As for 'unified' cooling core, I think the 'unified' part will be that it all runs from the same impeller. The triangular heatsink will actually be three heatsinks, surely.
That or we're going back to the days of the mini ps2 where one keeps a can of water on the top to stop it from stalling from overheating.



And modularity? lulz, sif. It's apple. They make consoles that do real computer work instead of pretending they're not real computers.












Summary:

Things I like about this:
The cool form factor in that everything is built around a central cooler. Pretty fucking cool if you never want to update anything.

The things I don't:
Everything else, including the issues the form factor brings, like everything being around a central cooler. And how un-expandable it is.

I've long wondered why 'normal' computer motherboards limit themselves into the same shitty form factor, and have continued to do so for the past 15+ years.
Shit gets cramped, hot & hard to manage. At least this kinda deals with that a bit with some lateral thinking. But it's not expansion modular unless you intend to add memory through externally attached swap space (read: an ssd via thunderbolt purely for faux-memory purposes).

Yeah there have been things like mini itx, but they all generally follow the same format. Slots one over the other with a CPU on the top, ram & storage connections off to the side. All in one big shitty rectangle.
What if we had the ability to lay our machines out how we wanted? Imagine how cool it would be to structure your own case mod however you wanted because it didn't really matter where the components were so long as they connected up in the end.
A bit impossible at the moment because wires are shit and slow, but how cool would that be?

And then how cool would it be to modularise the components of the motherboard? What if you could leave out the sound from the south bridge in favour of independent and superior solutions (dedicated sound card), etc?


Kinda like arduino but for performance PCs.
(also, fuckup pc versus mac, mac IS a pc, always has been, forever will be.)

[Xenolightning]

Except the 3 heat sources aren't in thermal contact with one another.

Where aren't they in contact with each other? In the closed loop water cooling circuit, or the "single extruded piece of aluminum"?
 

That image is grossly exaggerated, it is a common misconception that water heats up as it goes around the loop; it, in fact, does not. The temperature of the entire loop only fluctuates by a fraction of a degree, no matter where you measure it. This is due to  waters thermal properties and LOTSA PHYSX BRAH.

To iterate MCC's point:

There is very little real world difference between, attaching all components to a water loop, or all to one big heat-sink. A direct heat-sink of equivalent thermal dissipation to a radiator will result in the same temperatures assuming all other factors stay constant.

By attaching the components directly to the heat-sink they are merely reducing the "time" it takes from heat too go from a compnent to the heat-sink, metal -> metal conducts better than metal -> water -> metal. But the time it takes for heat transfer is irrelevant once the thermal dissipator reaches the critical heat transfer to air. This is where the difference will be seen, before you reach critical dissipation.




Also: Critical Dissipation. lul

[Bounty Hunter]

I was gonna reply to each of you but it's honestly not worth getting our knickers in a twist about.

It's just a dumb mac, and some boring heat sinks.

Playstation 4 and xbox are more interesting than this, seriously guys.

Although, discussion around pyro's idea of custom designing motherboards and PCB would be cool, except the poor bastard who has to lay all the track. We used the "auto-track" feature in altium pro a few weeks ago, it set up jumpers everywhere, though we only had 1 layer to work wit. I'm sure if you gave it 8 or 9 layers it'd do a half decent job. Your next problem would be noise, it's this whole other thing. And then you'd have muppets custom designing their mobo's with no CPU port or something....

[Pyromanik]

Yeah well I kinda envisioned more of an idea where you buy into a compatible system, rather than entirely designing your own Mobo. Chances are the companies who employ hundreds of eggheads can do a better job than most part time enthusiasts.

[private_hell]

Although, discussion around pyro's idea of custom designing motherboards and PCB would be cool, except the poor bastard who has to lay all the track. We used the "auto-track" feature in altium pro a few weeks ago, it set up jumpers everywhere, though we only had 1 layer to work wit. I'm sure if you gave it 8 or 9 layers it'd do a half decent job. Your next problem would be noise, it's this whole other thing. And then you'd have muppets custom designing their mobo's with no CPU port or something....

anyone who uses the default settings for auto route in altium should be taken out back and shot - even on 9 layer boards. your very very naughty bounty

default settings for auto route is like giving a 4 year old a colored marker and nice new painted wall

[Bounty Hunter]

anyone who uses the default settings for auto route in altium should be taken out back and shot - even on 9 layer boards. your very very naughty bounty

default settings for auto route is like giving a 4 year old a colored marker and nice new painted wall

no no no no dont get me wrong, we used the right settings, but we ran the auto-route wizard for a laugh....it was shit.

[Apostrophe Spacemonkey]

That image is grossly exaggerated, it is a common misconception that water heats up as it goes around the loop; it, in fact, does not. The temperature of the entire loop only fluctuates by a fraction of a degree, no matter where you measure it. This is due to  waters thermal properties and LOTSA PHYSX BRAH.

Got a source? Because I don't believe that.

Of course the water heats up. When you stick water next to something that's hotter, the water gets hotter. That's how my coffee is made.

[Retardobot]

That water is only in contact with each 'hot object' for a fraction of a second. It's being passed through, it doesn't hang around to heat up.

[Apostrophe Spacemonkey]

That water is only in contact with each 'hot object' for a fraction of a second. It's being passed through, it doesn't hang around to heat up.

Then where does the hotness go? The thermal energy must get transferred to the water, otherwise it would stay in the heat sink.

[swindle]

From a plumber who worked in large scale commercial - Xeno is correct.

[Apostrophe Spacemonkey]

From me burning my hand when I touch the radiator on my car, I don't think he is.

The water/fluid in the car cooling system starts out cold. Engine gets hot, fluid gets hot.

Why would it be different in the computer cooling setup?

[PrinceTuiTeka]

Have a read:

http://en.wikipedia.org/wiki/Latent_heat

BH can probably write a post or two on this explaining better than I can but :

TL;DR - Heat and temp are not the same. Temp can be a product of heat energy but not always.

[swindle]

From me burning my hand when I touch the radiator on my car, I don't think he is.

The water/fluid in the car cooling system starts out cold. Engine gets hot, fluid gets hot.

Why would it be different in the computer cooling setup?

Its not, you just have a significantly higher temperature threshold with a cars cooling system.

Like with the water loop, the entire loop heats up, but the parts in question don't produce nearly as much heat as an internal combustion engine.

[Retardobot]

Then where does the hotness go? The thermal energy must get transferred to the water, otherwise it would stay in the heat sink.

The heat transfers to the water, but because of the volume of the water and how quickly it moves through the loop (and being cooled again via the radiator) this heat is quickly dissipated. It doesn't allow it to heat up as much as you would think.

[Apostrophe Spacemonkey]

The heat transfers to the water, but because of the volume of the water and how quickly it moves through the loop (and being cooled again via the radiator) this heat is quickly dissipated. It doesn't allow it to heat up as much as you would think.

But your saying it does heat up? If only a little bit?

Because that's always been my point. Xeno said it doesn't heat up.

[Bounty Hunter]

Have a read:

http://en.wikipedia.org/wiki/Latent_heat

BH can probably write a post or two on this explaining better than I can but :

TL;DR - Heat and temp are not the same. Temp can be a product of heat energy but not always.

Latent heat is the energy required to change state? this is where you pump heat into the system but the temperature stays the same (because the system is changing state, for example, water changes from a fluid to a gas at 100 degrees Celsius but you can keep pushing heat in and the temperature will stay at 100 deg cel until the water is steam )

I think the assumption here is that we're working with a fluid which is not changing state, so either water below 100 deg/cel or oil.

[PrinceTuiTeka]

Hmm, yea sounds right. But heat can be latent without the change in phase right?

This physics is beyond me.

[Retardobot]

But your saying it does heat up? If only a little bit?

My original argument was in support of Xeno, that the temp. of the water changes only by a fraction, so yes. I believe you were arguing that the water should heat up considerably?

[Apostrophe Spacemonkey]

My original argument was in support of Xeno, that the temp. of the water changes only by a fraction, so yes. I believe you were arguing that the water should heat up considerably?

No sir.

I think he implied from my awesomely drawn diagram that the red colours was significantly hotter then the blue colour. I only meant to convey the transfer of thermal energy in a simplistic term, not indicate the absolute temperatures.

No harm done.

[Retardobot]

No sir.

wat?

It is a common misconception that water heats up as it goes around the loop; it, in fact, does not. The temperature of the entire loop only fluctuates by a fraction of a degree, no matter where you measure it. This is due to  waters thermal properties and LOTSA PHYSX BRAH.

Got a source? Because I don't believe that.

Of course the water heats up. When you stick water next to something that's hotter, the water gets hotter. That's how my coffee is made.

[swindle]

It doesn't matter how many points you deal with in a water loop, and how hot they are, the whole loop will become the same temperature. Loop cooled efficiently, it will be a low temp

[Bounty Hunter]

We all understand that the temperature of the water will depend on the speed at which it passes the element right?

[Retardobot]

I'm having red beans for tea.

[Growler]

Im having sex and bacon

[The Demon Lord]

This is clearly a giant Butt Plug so that Apple Fanbois can get even more intimate with their Macs....