Scarlet-Tech
I have plenty of acetone, but it never went below -72 when I tried last time. Been too busy to try again.

You are Correct because -72C equals about -100F!
 
In my Previous pic of my Digital Thermometer it is set for reading temps at Fahrenheit, So the reading it shows at -99.4F is really about -72C.
 
That's Why when you see Benching Rigs using LN2 at -196C it is REALLY at -321F!

tucker147
Is it fine to use dielectric grease to coat the board (and cover the ram & fill the slots)? I ran out of usable LET when I was trying to add more layers and I have a bunch of grease around.

You can use whatever type of insulation methods that you know will work.
 
Others will tell you There are Many ways to condensate-proof MB's, Graphics Cards, etc...and they DO work well, But Since 2008, I have used ONLY Petroleum Jelly on All my Sub-Zero Overclocking/Benching Rigs Now and in the Past.
 
My Method is using Petroleum Jelly and apply it on Both Sides of the MB Area you think will get condensation and IN the CPU Socket, Then use a Heat Gun/Blow Dryer to cause the PJ to become semi-Liquid and FLOW into and Underneath all the areas of the MB, Especially under the Memory Slots, CPU Socket and then AFTER I install the CPU and lock it in place, I heat it again to ensure good electrical contact between them.
 
Look at my Post #10625 on this Link:
 
http://forums.evga.com/Get-It-Colder-II-m508-p355.aspx

tucker147
Thanks, I've got it all finished up now. Though I didn't use as much grease in the socket as you did in the pics, but I had a little scare on boot it kept giving me post code FF. I cleaned up the cpu and socket a bit and it ran fine again.
I couldn't get this stuff to melt at all with a hairdryer but I don't think I'll run into any condensation problems, I completely covered the top half of the board front and back with the grease.

Yeah, Dielectric Grease won't melt and flow like PJ does, But still Glad you got everything worked out OK.
You'll more than likely be able to spot any condensation problem areas that may show up when you start going sub-zero with the DICE/Acetone Cooling.

Congrats, I assume that BIOS reading is still with just Ice/water, But at least Now that you had a "taste" of sub-zero cooling you will never look back!
 
Can't wait until you try a small amount of DICE for the first time. You could even use some 91% rubbing alcohol instead of Acetone just to start off, But the sub-zero temps will be somewhat Higher (Around -51C/59F).
 

Alright, I've gotten all my S775 stuff yesterday, did a bit of testing once I got it set up.
P4 631 @ 4.5GHz 1.43V This particular chip can do 5GHz easy on air apparently so I'll keep trying for that to practice.
 
System, P5Q-E with a coat of LET and volt modded & 2x 512MB Crucial Ballistix 667MHz 3-3-3-12 @ 2.2V.

Volt mod controls,

CPU's, some Celeron's, Pentium's, Xeon's as well as a E5540 ES and a Cele D 352 that did 7GHz on another mobo, these should last me a very long time.


 
I'll be getting some DICE this weekend hopefully. 
 

hallowen
Looking Good, I'd say it Definitely looks like you have Plenty of CPU's for Testing!
 
I've always enjoyed Raising the Voltage and Pouring on the Cold to CPU's just to see what Limits they can endure before Meltdown occurs.

Well, at least I won't get bored any time in the near future. 

Tell me more about that volt mod daughter board??  I like it.
 
Funny, when I see that 5160 CPU I think of spring steel now. 

Hi everyone.. Running some benches and tests right MEOW!!!!
 
This is my water temps (ignore that it is set to #1 and #4, the connectors are too big to fit in #2 and #3)  This is using Ambient air temps btw.  11c-16c under full benching load.
 

 
Ignore the mess on the desk, this is how i am getting it colder today (up to 1599mhz on the GPU so far and 4.64 on the CPU):
 

 
I am running into a strange error* (if that is what it can be called) on my 980ti Classies.  When I hit 1599 max boost, no matter how much I overclock, GPU-z is only reading up to 1599, not past.  The card is stable up to +195, but once I hit +156 (the point it hits 1599 max) it won't read past that.  Any idea what could be causing this?
 
 

johnksss
Did you ever figure this out?

Not yet. The temp went crazy and jumped from -2c in the room to 22c in 2 days.. it's like mid spring here right now, lol.

Waiting for another cold day to try again.

Leaving town (the country actually) in a couple of days. May have to wait until I get back unfortunately.

I have already reapplied the TIM. I tried the star method instead of spreading it, and the temps were not as good as I am used to with the loop being so cold. I redid everything, and now the temps are lower and both are nearly equal. I think the TIM was allowing a part of the core to get hot and causing failures.

Scarlet-Tech

johnksss
Did you ever figure this out?

Not yet. The temp went crazy and jumped from -2c in the room to 22c in 2 days.. it's like mid spring here right now, lol.

Waiting for another cold day to try again.

Leaving town (the country actually) in a couple of days. May have to wait until I get back unfortunately.

I have already reapplied the TIM. I tried the star method instead of spreading it, and the temps were not as good as I am used to with the loop being so cold. I redid everything, and now the temps are lower and both are nearly equal. I think the TIM was allowing a part of the core to get hot and causing failures.

Hummm, I was able to pull this off on air....
I was testing under windows 7 at the time.

 
So it has me wondering what is going on. Do you have any single runs listed anywhere? I want to check to see if those match up with any of mine. Thanks.
 

http://www.3dmark.com/fs/7726269
 
I don't have an evbot. You know anyone trying to sell one??
 
http://www.3dmark.com/compare/fs/7726269/fs/7748638
Why does your clocks show 135 mhz and 162 mhz for memory?

Ok, thanks.
And at stock, it should be 1430 on core when running optimal.

You can see mine... It's 74 Asic.
In the laptop cards that Asic score means a higher or lower voltage and a higher or lower memory overclock. I have a 80 and a 70. the 80 card can do 1500+ memory over clock, while the 70 card can only do 1430 to 1450 stable..... Not sure if that is the case with desktop cards.
 
 
Edit:
Memory testing...
1430C/2103M
http://www.3dmark.com/fs/7724447

Well, it's been awhile since I made any posts or replies on the Forums due to anticipation on the New Pascal offerings to become available, But Alas, I see where it has been confirmed by EVGATech_ChrisB that the First offerings will NOT be compatible with 3 or 4-way SLI (Only 2-way), So looks like Unfortunately I'll have to be waiting a little longer for something to feed my 4-way Habit.
 
Looks like in the meantime, I'll maybe have to concentrate on trying to obtain an Intel 6960X.

I have some news.. My phase Change is in for repairs and upgrades right now, so hopefully I will get that back in a few weeks, and I can start running the test bench a lot more often without needing to go get DICE.