Ivy Bridge hits the streets--post your results here

[SAAB340 28]

then I'll image and I can run some tests again. Don't know how long it will take this time, probably a couple weeks.

 

No probs. I should have noticed the csvs earlier when you posted them.

I hope you get your system stable soon so you can enjoy using it. It seems like a nice setup you got there.

[TechguyMaxC 516]

It's stable now, I didn't elaborate enough earlier. I'm going to be swapping out my Corsair H100 for a custom water cooling solution in the effort to reduce CPU temps further as well as overclock more. Going to cool the 680 as well, temps there are supposed to get down to around 40 degrees C so that will be nice.

 

After that I need to reconfigure FSX for FSXMark 11 testing and make an image that I can revert to at any time for further testing. Then I can install all my add-ons and configure flight controls and tweak the config file. I really want to get to the flying so I may skip the testing phase for now and just go to the configuring phase, then make another image once I have it fully configured. Then I'll be able to switch between benchmark-ready image and flight-ready image on a whim.

[TechguyMaxC 516]

Finally got around to installing my custom water cooling loop and I am very happy with the results. With the old H100 and IC Diamond thermal paste between CPU core & IHS and IHS & H100 waterblock base I was hitting temps of 104 on the hottest core in Intel Burn Test. Now, with Coolaboratory Liquid Ultra liquid metal thermal paste between CPU core & IHS and IHS & XSPC waterblock, using an XSPC RX360 radiator, Alphacool VPP655 pump, 7/16"/5/8" tubing, XPSC Raystorm CPU waterblock, and XSPC Razor GPU waterblock with the pump at the lowest speed for initial testing I have achieved the following results:

 

Intel Burn Test maximum temperature on hottest core: 72C

 

That is a drop of 32C.

 

That is astounding, completely blowing away my expectations. I believe this puts the nail in the coffin on the Ivy Bridge thermal paste debate. Better paste + better cooling yields significantly lower temperatures. If Ivy Bridge suffered some kind of process limitation or thermal density problem this simply would not be possible.

 

Now testing with fixed voltage of 1.32V and 4.8GHz, will report results when I have them.

 

Here's a quick pic I took from my iphone:

[Kosta 122]

Finally someone else also going a full watercooling loop. Welcome to the club, Max!

[Dirk98 354]

Techguy, congrats! As I've mentioned the result is inspiring enough to burn a couple of 3770k's.

 

A few questions here, if you don't mind:

 

1) Was it hard to remove the IHS with a razor blade? Could you comment and give some more details on how you managed this. Was it something like in this video:

 

http://www.youtube.com/watch?v=jXaTCm4Uguk

 

2) How did you remove the previous thermal paste from the core? How clean you left it before applying Liquid Ultra?

 

3) How thick was the layer of Coolaboratory Liquid Ultra you put on the CPU core, and how hard did you spread it? What did you use for spreading it?

 

4) How did you put back the IHS and seal it?

 

Thanks,

Dirk.

[TechguyMaxC 516]

Dirk, I have to head out to church now so I don't have time to answer your questions but wanted to let you know I saw your post and will respond as soon as I can.

[Dirk98 354]

Take all the time you need, of course.

 

Thanks for sharing.

 

Dirk.

[TechguyMaxC 516]

Techguy, congrats! As I've mentioned the result is inspiring enough to burn a couple of 3770k's.

 

A few questions here, if you don't mind:

 

1) Was it hard to remove the IHS with a razor blade? Could you comment and give some more details on how you managed this. Was it something like in this video:

 

http://www.youtube.com/watch?v=jXaTCm4Uguk

 

2) How did you remove the previous thermal paste from the core? How clean you left it before applying Liquid Ultra?

 

3) How thick was the layer of Coolaboratory Liquid Ultra you put on the CPU core, and how hard did you spread it? What did you use for spreading it?

 

4) How did you put back the IHS and seal it?

 

Thanks,

Dirk.

 

Dirk,

I've got time to answer your questions now, I hope it all makes sense.

 

1) My removal process was much like that shown in the video, though I used a slightly different tool, something like a paint scraper. The other two Ivy Bridge CPUs I've de-lidded for friends have been with bare razor blades though, and I find this to be the best method thus far because you can wedge the blade in and just push with your thumb until it goes through. All told it is perhaps a 5 minute procedure. I also recommend the removal of the excess epoxy material from the CPU package and the underside of the IHS as this allows the IHS to sit lower on the package (and thus have less space between the CPU core and the IHS necessitating the use of less thermal paste).

 

2) I just used rubbing alcohol and paper towels until all traces of the previous paste were gone.

 

3) Coolaboratory Liquid Ultra is a very thin thermal interface material, it truly is a liquid metal so it spreads incredibly thin. I used some small paint brushes to spread it (they supply a couple small brushes, I got a couple larger ones to make it go faster). You don't want to press hard on it at all, use very light brush strokes to spread.

 

4) The IHS just goes back on over the CPU core, there's a notch on one side that helps you orient it correctly, it simply lays on top and the adhesion of the thermal interface will be sufficient to hold it in place until you put it back in the socket and clamp the socket retention bracket back in place, though I recommend laying the motherboard flat when you do this just in case, no sense working against gravity. I did a test run to see how good the contact was and pulled the IHS back off - there were definitely some light spots so I just filled in the areas where there was no liquid metal on the underside of the IHS and then put the IHS back in place. Then I spread the Liquid Ultra on the topside of the IHS and mounted the water block and that was it.

 

Testing at 4.9GHz now, looks like it's going to need a lot of Vcore beyond 4.8GHz as I'm at 1.415 and still testing for stability. I'm sure 5GHz is within reach, just hope I don't have to get too close to the 1.55V limit.

[Dirk98 354]

TechguyMaxC, please carry on your experiment with IB, let us know about the progress. Very interesting. Thanks for taking the time to share your method and results.

 

Cheers,

Dirk.

[idahosurge 2]

It is a couple of months old, but for those how are interested here is a great benchmark comparison of:

 

3570K vs. 2600K, both at 3.4GHz with HT off,

3770K vs. 2600K, both at 4.5GHz with HT off.

 

All benchmarks were ran on the same system with only the CPU changed.

 

Even though none of these are FSX benchmarks it is interesting and IB won everyone.

 

http://www.xtremesystems.org/forums/showthread.php?280352-Large-Ivy-Bridge-3570K-vs-Sandy-Bridge-comparison.

[westman 257]

HI

I normally dont use burn test.

but i do tests.

Here is a Burntest on a 3770K "no mods on IHS" and 12C on IHS