Direct Die Liquid Cooling... (not Water)
 

1 degree C over Ambient at Idle


33C after 6 hours of Prime 95

at idle my temps are 24C mobo/ambient - 25C cpu

Under load I get mobo 27C - cpu 33C

Amd Athlon 2600 Barton Core (stock 1.91) running at 2ghz

running a vcore of 1.85V (max for this mobo)


Best I can get is 2.1 GHZ...

Been O/c for a couple of days now, witrh lots of different configs, I even threw in some PC4000 from my main rig and only got an extra 10Mhz out of it, I think the procc is at it's limit...

If not water, then what? Certainly not peanut butter...?

Peanut butter cooling would be humorous. Saw this on [H].

I'd imagine that peanut butter would have a high thermal capacity. Its viscosity might pose a problem though :p

Using flourient?

I never thought of peanut butter... Maybe I will try that instead of AS5 on my next system...


Anyway... The fluid if HydroFlouroEther

Why is it that you had no problem telling everyone here what it was, but you pretended it was some big corporate secret on PCPerspective forums?

UMM... because at one point it was....

The system was going to be Patented because no one ever did it before the way I did it, I even suprised the Engineers at 3m with it...

but now I don't give a sh1t about it...

I already put $700+ into this and don't have anymore money to spare to go as far as I wanted to go with it.... and Patents etc, are going to cost way to much, so why not just say f*uck it, and let someone else have the knowledge and maybe they can do something with it................

If you don't know what is going on, then don't start with me...

and on PC perspective I told them as well... the next post after the guy asked me what it was... So where do you come off with this?


Damn nice welcome i got here... Thanks!

How is the direct die block mounted? How is it that it does not leak?

It's mounted with pressure by the Screws/Posts, I have a piece of Plexi on back of board as well..

and the seal between the plexi and cpu is closed cell neoprene..

We tested this at AVT last year, you will have long term issues with some fluids.
Current testing another fluid for three months now, too early for results.
Can confirm you will get improved results if you work with design.

As for something "new", no, we understand someone before any of us had a product on the market a few years back, no idea if they patent or what fee will be, sorry.

You should also be aware, direct on processor die cooling channels are being tested right now.... not even "inside" info any more, and very well patented.

Hope this info helps before your $$ go into developement.

For home guys, this should be your next area of research, it can work very well as Draco shows. Long term is still under research.

Required pump pressures can be low. What will be required for built in channels is still up in the air, seen research for both high and low pressure, with about equal results.

Dave

Hey Draco!

That is NICE dude!

I'm really impressed!

Yeah the block design was just a prototype, I have 4 more to work with, I just wanted to see if it would be possible to get it up and running with no leaks, that the seals I am using will work etc. and maybe get an idea of just what this stuff can do if applied correctly

I had it in my setup using the mase 4 copper block and the results weren't that great, they are 100 times better with the direct die setup..

as for low pump pressures, I can atest to that, as I slow my pump down, to reduce the bubbles, as this fluid actually is thinner than water, and produces it's own bubbles for some odd reason... even the techs/engineers at 3m can't figure it out, except maybe some sort of phase change is happening in the pump and creating the bubbles, or some otehr things going on that I/we don't quite understand yet...

but I know when I slow it down the bubbles go away..

as for $$ into development, I don't think it will happen unless I get someone to invest (which is doubtful)

It was a cool project to see if it could be done, to do something NEW and Exciting inthe Private sector!!

Bubbles are from pump impeller, if we are talking same fluid as I suspect we are.
(can not openly state)

We found it stops with a 70 deg end tip angle, but depends on velocity.

Also be aware, we are not 100% sure it was an exposure issue that caused a failure, that is why we are re-testing.

Processor could just have died due to natural causes :)

Yeah that's what we figured that the impeller was causing the bubbles, that's why the bubbles dissappear when I slow it down...

As for Long Term research, I know of systems that have had Direct Die Cooling for 4+ years and are still working fine with this particular fluid...

Good to know, thanks for sharing.

But maybe a moot issue if processors come with built in cooling channels.
Even so, real interesting field to be in right now :)