Old water blocks on modern processors

[sunrunner20]

Well, looking at some reviews (primarily http://www.systemcooling.com/swiftech_apogee-01.html) I've come to the conclusion that since most modern water blocks are focused on small die area cooling that I have gained an advantage in that my 6002's general cooling will actually perform better than, an apogee or swiftech storm.



now... about that mounting bracket...

 

[Nandro]

I may need to buy the $10 775 mount for mine so I can use that. kinda useless without the mount unless someone wants to trade for an amd mount.

 

[nikhsub1]

Well, looking at some reviews (primarily http://www.systemcooling.com/swiftech_apogee-01.html) I've come to the conclusion that since most modern water blocks are focused on small die area cooling that I have gained an advantage in that my 6002's general cooling will actually perform better than, an apogee or swiftech storm.

And you would actually be 100% incorrect. Did you not see the graph I posted above?

 

[sunrunner20]

Forgive me if I trust a known testing platform(systemcooling's) more than an unknown(yours).

 

[SolidxSnake]

How about the WhiteWater? Does that cool near the 6002? Worse, better? Anyone want to test their WW on a quad?

 

[nikhsub1]

Forgive me if I trust a known testing platform(systemcooling's) more than an unknown(yours).

yeah i used to be like you but die sims dont translate to IHS cpu's AT ALL. Look at his CPU testing...

 

[Voodoo Rufus]

The results are so close (well within the margin of testing error) there are no clear “winners” (or losers). All three waterblocks performed very well. This shows that under normal conditions the Apogee can perform on par with the MCW6002 and Storm – just like Swiftech claims.

It ain't perfect either. For some reason there is more variation on Nikh's testbed.

 

[nikhsub1]

It ain't perfect either. For some reason there is more variation on Nikh's testbed.

Yeah, more heat plus a true on die reading of core temps...

 

[sunrunner20]

Just realized... my mcw6002 won't even work on a lga775 system. It doesn't have a stepped base so I am kind of screwed... Looks like I'm out of the water crowd for a while...

Nik, do you have stats for unbowed versions of the block?

 

[Moto7451]

Thats why God gave us the Dremel tool .

Are the quad cores soldered to their IHS? If not you could always remove the IHS which should help your 6002 seeing as it wasn't designed to deal with one.

 

[Mustanley]

Got my goods today, I'll probably use the stock HSF while I verify all the components are working (bought a refurb P5K Premium for $175) then get my WC gear installed later tonight or tomorrow as time permits.

 

[SolidxSnake]

Thats why God gave us the Dremel tool .

Are the quad cores soldered to their IHS? If not you could always remove the IHS which should help your 6002 seeing as it wasn't designed to deal with one.

I'm pretty sure they are but you can remove the IHS's anyway.

 

[Moto7451]

Yeah, I was just thinking that I have a hard enough time with just a razor blade and I had to kill 3 helpless semprons before getting it right (well took all their caps off, they still run, just can't OC) . Are the soldered CPUs also "glued" to the IHS?

 

[SolidxSnake]

Yeah, I was just thinking that I have a hard enough time with just a razor blade and I had to kill 3 helpless semprons before getting it right (well took all their caps off, they still run, just can't OC) . Are the soldered CPUs also "glued" to the IHS?

Well, the IHS is glued to the CPU Substrate with the black sealant normal IHS'd CPUs are sealed with. The IHS is most likely also soldered to the die of the CPU on quads, but I can't confirm it.

If I get a C2D I am SO delidding my P4 631 (which isn't complete yet)

 

[Moto7451]

Yeah, seems like I'd need to destroy more than a few intel products to get that one right .

 

[Mustanley]

Decided to lap the Maze4 in preparation. It's got some dings and nicks from a few dozen mountings over the last 4 years.

Still trying to work some kinks out the system, but I have gotten this Q6600 stable on the stock cooler at 3.2GHz. After I figure out why the multi keeps dropping in windows, and make it stick, I'll be ready to assemble everything the loop.

*edit* Ok, it's throttling from heat. Finally got core temp working in vista64 and saw it hit 100C while priming. Must not have the hsf seated properly. Might as well go ahead and mount the waterblock

 

[Mustanley]

It's been a long night, but here were are at 3.4GHz. Took forever to get the WC loop installed as there always seem to be unforseen problems.

Anyway, I jumped straight to 3.4GHz with the Vcore on auto.

Tomorrow I'll see how far I can push this chip.

Room temp is 72F/22C

 

[Nandro]

good job. the water cooling should help immensely on that cpu. Granted some have gotten good clocks with air vut the average is going to be lower. I may buy The coolit freezone elite when it comes out. My wife says I destroy too much stuff building my own, and now she's making me suffer for it (really we do have alot of bills) But keep going. Whats your max temp after water load for say 2 hours?

 

[Mustanley]

I let it prime on small FFTs for 30 minutes and never broke 59C on the hottest core. Then I switched to SMP Folding which runs about 5C cooler.

Downloaded CPU-Z 1.41 and it says my Vcore is 1.384 (loaded), but I've got voltage set on auto in the bios and I'm thinking that I might able to lower it and maintain stability. First though I'm gonna see how high I can clock it keeping temps under 75C or so.

Honestly, I'm glad I didn't bother getting a whole new WB. Sure I might have been able to lower temps by a couple of degrees, but in the end I doubt if it will limit my OC one iota.

 

[sunrunner20]

I Doubt a dremel would help me with my 6002...