Sapphire Cooling Technology Rocks

More pics and info here: http://www.xbitlabs.com/articles/editorial/display/e3-2005.html

The cooler seems effective enough.

 

Seeing these pictures i doubt that a conventional cooling solution of this size wouldn't allow such clock frequencies.

 

Hehe, yeah, i guess so. But it's pretty sweet to get such good cooling out of the box though.
I have an AC Silencer 5 on my X850XT PE, and 650 MHz is out of reach. I'm pretty close though..

 

colloids are bascially what forms when you have two little chunks bump into eachother and they form a bigger chunk DW, eventually you will just get a solid lump, unless you do something to make sure the the equilibrium favors them staying dispersed which I assume is something related to what mufu is talking about.

That is how they treat the water you drink usually is mess with the equilibrium to favor the formation of colloids so that all the nasty crap lumps together and sediments out.

 

Sapphire have signed a six month exclusive with the company (similar to what they did with Zalman on the original Ultimate editions) so they have a six month lead time before any other VGA solutions can pick up on it.

Onething that is quite neat is the fact the liqid is ferromagnetic metal you don't actually need a "pump" in the traditional sense, just a small electromagnet that pushes the liquid around. The solutions they have at the moment are early engineering samples and they are expecting to only use one fan on the final versions and this may not be on all the time dependant on workload; they will also plug it through TRIXX's profiling so you can se up different types of operation dependant on what's being run.

 

Am I the only one who doesn't like the idea of small electromagnets down by all my system components? (though I do like the concept of the cooler)

They'd best be awefully small, or else shielded.
I can picture the complaints already about noise/crackling/whatever from sound systems during processing (gaming, video, whatever).

 

I wonder how much weight this will add to the card

 

A colloid is a suspension of particulate matter which tends to not settle or precipitate out of solution.

That's flocculation & would be useless in the case of a ferrofluid for cooling.

 

stevem go the nomenclature right. Small clarification: in "particulate matter", the particles shouldn't be interpreted as necessarily being "solid", milk is a colloid for instance. Besides, what they call liquid metal seems to be just that - liquid metal. No colloidal phenomenon involved.

Be that as it may, I can't see this cooler as having any strong advantage. It may (or may not) have an advantage vs pure copper as far as heat transfer goes, i.e. it may be somewhat better at dealing with the problem of local hot spots on the chip. If so, that is good. But that doesn't mean diddly squat as far as heat dissipation goes. If you don't actually dissipate the heat from the cooler and into the surroundings, you just end up with a hot cooler. And as far as the dissipation problem goes, this technology offers nothing, so it may well be that what limits the performance of this cooler is how well you manage airflow across cooling fins, just as with all the other coolers out there. I do hope they make a two slot solution in order to enable dumping the hot air outside the case. If they don't, I can't see this solution doing anything better than a VF-700Cu for instance.

It may be that in the future, taming on die hot-spots will be the major challenge, if die sizes shrink and total heat output stays the same or is reduced (, lowering the significance of the dissipation problem). But it isn't the case right now, and I doubt it will be in the near future either for PC GPUs.

I'm sorry to be a bit of a wet blanket for what looks like an interesting cooler, but I feel that right now it may be a good solution to the wrong problem.

 

Further clarification. Unless we're talking about filtered, homogenized milk, it isn't always the best example of a colloid due to settling of various lipids & solids. No inference that the liquid metal in question is a ferrofluid. However, is an alloy of two liquid metals at room temp, for which there's uniform distibution of metals within the alloy & hence no aggregation a colloid?

 

hmmmm.... and here I was thinking a colloid was something you left in the bowl when experiencing particular troubles!

 

Wouldn't say so, no. I'd say the concept of colloid implies a non-precipitating suspension, rather than a mixture. At what precise proportions a suspension turns into a mixture is the kind of question I tend to avoid. Science is the extension of craftmanship, not philosophy.

 

Yeah I know that was my point they would do something to push the equilibrium the other way to keep the particles in suspension.

 

It makes me a bit edgy, but I've also went against conventional wisdom and have been using a magnetized screwdriver on PCs for about 5 years now and have suffered no ill effects from it.

Little magnetic fields are ok, but ya gotta keep 'em small.

 

This reminds me of one of my wife's former TV stations. . .the IT department was having trouble with the integrity of their backups. Their tapes were always coming up dodgy.

Turned out on the other side of the wall from the tape archive was the video department. . .with a nice size tape eraser against the shared wall with IT that folks would slap their pro beta tapes down to be erased for reuse.

Oy.

 

Venture capital funding.

 

Hehe, good call.

 

This is DOA now.

 

OMG Ty, Ty, Ty........you bumped a FIVE month old thread just to say it's a dead topic?!?!?!