Dominance RAM Pot & LN2 // Corsair Dominator GTX2 // 13 Hour Cadda 6 Party

[3oh6]

Well it is about damn time I posted some results with this little pot after talkin shyt for so long about the insulation. Like the good little reviewer that I am, I completely forgot to take photos of the pot before I started insulating it . Needless to say, it is a block of aluminum and if you really want to see it naked you freak, head over to K|ngp|nCooling.com as Vince has a photo posted there. There should be some internal shots in this thread or my insulation guide thread.

As mentioned, I have already done a detailed insulation thread so I won't go into details here. The Dominance Memory Cooler insulation guide can be found at this thread here. I prepared insulation in hopes of being able to run extended sessions for testing memory...today I went over the top in proving that the insulation can handle extended sessions. Let's just call this post a memory clocking bender. I simply wanted to relax, have some fun, and test sub-zero memory for an exorbitant amount of time...mission accomplished, and got a couple fun results too boot

***\start big caveat to the results/*** I completely forgot I was using a development BIOS from ASUS for this testing until someone inquired about the version in the screen shots. You will have to call me Pro Jr./Hicookie Jr. here as I am using a BIOS that potentially inflates 32M performance over that of the retail BIOS's. In my quick testing of this BIOS i found it gained about .3 ~.4 over the previous BIOS I was using (0402), but I didn't exhaustively test it. Some are reporting higher performance gains but I think it is user error on their part

Either way, here is a link to the 0046 BIOS. Don't say I never gave you whining brats anything. Use at your own discretion, if you bork a BIOS, you have a second on board...bork two and you don't deserve to have a board and I am in no way responsible for your special olympic qualifying behaviour.

I am not propping up the R3E's 32M performance on ASUS's behalf in this thread, it just happens to be the board i prepped for sub-zero...that is all. I am being paid handsomely by Vince (think in the hundreds of thousands of $$$'s) to promote his $45 RAM cooling pot. Don't read anymore into it you illiterate forum monkey's.
***/end big caveat to the results\***

Now enough of the bullshyt, on to some photo's and screen shots.

Click for full size...

From left to right, we found ice forming on the front edge of the board at Vince's during our initial run. I remedied this by extending the eraser right to the edge of the board, and mated it to a couple layers of Armaflex tape. This should seal up this edge nicely.

With the base insulation and evap mounted, all that is left is to put the pot on. It was at this point that I realized just how sexy this insulation setup is. I knew it was going to be a long night as I vowed to bench until the LN2 was gone if the rig wanted to.

The last photo above simply shows me warming the system up before turning on the phase...everything registered correctly (AKA all three slots were up and running Patrick...yes, I checked before putting the insulation on this time ).

Click for full size...

These are just some random photos taken during the 13 hour session. For the first couple hours I kept pot temp at -25C'ish while fighting 7-7-6 clocks. I really didn't get shyt for clocks at CL7...which was also the case at Vince's.

The last photo is how the last 8-9 hours of the session were ran. Usually at 1.85v and pot temp at -50C ~ -60C. This is when I switched to cadda 6 and life was good.

Hardware used:
RAM:         Corsair Dominator GTX2 (reviewed @ Hardware Canucks)     
MB:          ASUS "R3E 95" (Developmental BIOS 0046)
CPU:         Intel i7 980x ES (1.425v VTT limit :( ) // Intel i7 975 ES (2:12 ratio is dead :( )
CPU Cooling: Chilly 1 single stage (tuned by Ruffus)
GPU:         HD5870 w/ghetto HR05-RevA
NB Cooling:  Noctua NC-U6
SB Cooling:  Enzotech SLF-1
PWM Cooling: Swiftech MC14's
PSU:         Corsair HX1000W
HD:          Seagate SATAII 80GB 8MB NCQ
OS:          Windows 2K3 "super wicked awesome edition"

Ambient Temperature: 22-23C

Click for full size...

I have already had requests...so here is your fog, frost, and outright snow shots

Basically, I am a loser. Addicted to the 32M LCC. It has been too long since I put any effort into anything so this session was dedicated to the sub 7min category. The goal was to max out the memory to get the best frequency I could. After my issues with 7-7-6, it turned into a Cadda 6 party...enjoy

5051MHz // 1010MHz 6-6-5 // 6m 59.516
Click for full size...

I might as well not piss around and just show the good stuff. This is the best I could pull off. With the CPU multi lowered, and BCLK raised I managed a 7m 00.149 @ 5041MHz...but the memory was tapped at that point and the last .149 was not to be had. The uncore limit of 4.2GHz'ish is the biggest thing holding this setup back. The Gulftown likes the uncore and this chip just can't supply it. Still pretty fun to run uber tight timings with no uncore and compete.

4992MHz // 1040MHz 6-6-5 // 1.85v
Click for full size...

There was just no way I could get the extra couple BCLK to get sub 7m with this CPU multi. Going higher in voltage didn't help get the extra clocks which leads me to believe either the sticks are just done there, or the IMC can't handle more. So call this the max 6-6-5 clocks...at 'only' 1.85v and -50C pot temp.

3998MHz // 1052MHz 6-6-5 // 1.88v
Click for full size...

This is a small update on 6-6-5 clocks with a 975 for the CPU. The secondary timings are stupid loose but I did manage to get this to run once. These modules don't like 1.90v+ and 1050MHz wouldn't run with 1.85v so I met them half way and squeezed a run out. Unfortunately this CPU won't run 5040MHz for a sub 7min run


3998MHz // 1052MHz 6-7-6 // 1.81v
Click for full size...

Again, nothing more would get higher clocks on the memory leading me to believe it is an IMC limit. The 1.425v VTT that this chip is limited to really appears to be hindering the memory overclocks...even at cadda 6. Again max frequency achieved here was 1052MHz 6-7-6 @ 1.81v -50C pot temp.

Overall the 13 hours I benched were quite informative, but unfortunately this IMC ran out of gas at possibly even 6-6-5, certainly at 6-7-6/7-7-6. I have another kit of Dominator GT 2000C7's that I might try with this setup just to confirm my speculations on the IMC.

As for the Dominance memory cooling pot, how can I complain? It is super easy to maintain temps within a tight range, and can easily handle extended bench sessions. Vince has done a bang up job with this pot. And for only $45! It definitely gets my . Here are some photos of tear down.

Click for full size...

As I mentioned, this insulation job is perfect. There is no moisture anywhere, and no vaseline in the slots is required. Even an extremely long 13 hour session with pot temps at -50C for the last 8 hours couldn't penetrate this insulation method which left the DIMM slots bone dry

[Brian y.]

Nice job Jody

[G H Z]

Beautiful setup, it was only a matter of time till we saw some real 'dominance' out of it

[Gautam]

Damn, new settings for performance wr.

[eva2000]

bloody nice 13hr session - dam now that's dedication

surprised the uncore frequency on that 980x is so limited

[k|ngp|n]

Of course Im not biased or anything , but sweet review. I saw him preparing his mem insulation in person. Perfect conainment of the cold zone. If you spend the time, you wont do the time.
Like someone else said, you could float down the river with that insulation job man

[Gautam]

Yeah, he's really meticulous.

[Perica_Barii]

One word : Perfect ......

[dinos22]

nice work mate

[systemviper]

great effort, thanks for the post.

I have what might be a bonehead question, can't get Ln2 SO how do you think that mem pot
would work with dice, to shallow? or something you might point out....
not looking to bench for long periods of time but for 45.00
it might be fun to play with.....

thanks in advance...

[dinos22]

great effort, thanks for the post.

I have what might be a bonehead question, can't get Ln2 SO how do you think that mem pot
would work with dice, to shallow? or something you might point out....
not looking to bench for long periods of time but for 45.00
it might be fun to play with.....

thanks in advance...

mate it would work the same just pack the dice without acetone and watch the temp probe

you honestly won't need more than -30C on the Hyper IC sticks as pro and I have done a lot of RAM subzero benching last year for F1OC

[3oh6]

thanks guys, i have been waiting to be able to have a memory clocking session like that for far too long

and yeah, Dinos is bang on...temps don't need to be low. dice in rice or pellet form would be ideal as it would prolly keep the pot temp right at -30C~-40C. keep in mind there is absolutely no load swing or anything like that, it is just a matter of keeping pot temp right where you want it.

to be honest, i wanted to do this session with dry ice. i figured it would be easier and less work. pouring LN2 was a PITA because you could only hit the pot with 10-12 drops of LN2 or the temp would swing too much...but you had to hit the pot every 20~30 seconds. made for a very very tiring session.

bloody nice 13hr session - dam now that's dedication

surprised the uncore frequency on that 980x is so limited

dedication or stupidity...the jury is out on that one. and yeah, low VTT limit of 1.425v i am sure is the deciding factor in the low uncore of this chip. but keep in mind, temps are only -30C at coldest at the evap so the uncore isn't "terrible", but still not good by any stretch.

[systemviper]

top notch guys ,,,, thanks

[dinos22]

dedication or stupidity...the jury is out on that one. and yeah, low VTT limit of 1.425v i am sure is the deciding factor in the low uncore of this chip. but keep in mind, temps are only -30C at coldest at the evap so the uncore isn't "terrible", but still not good by any stretch.

yes it's most likely the VTT volts. The temps you are running should not stop you doing 4.8GHz uncore if the CPU/board/bios/RAM combo is capable of it.

i know others that have chips that absolutely hate VTT so it might be just bad luck with that chip

[Utnorris]

WOW!!!!

Nice job!!

[chispy]

As always very informative and rigorous througth review wich excells above all others , loving the Ram Pot . Now Jody go get some well deserved sleep

[3oh6]

thanks ladies

yes it's most likely the VTT volts. The temps you are running should not stop you doing 4.8GHz uncore if the CPU/board/bios/RAM combo is capable of it.

i know others that have chips that absolutely hate VTT so it might be just bad luck with that chip

yeah, there are a lot of chips that are seeing VTT limits on them. but i have also seen a lot of chips that uncore scales with cold...this one in particular. under LN2 i can get uncore up to 4.4GHz range, and many others will attest to this as well.

looks like i need a cascade for memory testing, get a plate made to mount the single stage on the Dominator sticks...and test memory all day without getting up

threw the sticks in today back on ambient to check and see how they were doing...not a single sign of any issues. they maxed out at 1040MHz 7-7-6 @ 1.65v in testing on the day i got them, and i ran this last night at 1.65v on the DMM...



so all is good in the hood

[Hondacity]

the no swing....is because the sticks themselves are just 1-2w

too bad the imc is holding you back...

still i like the no frost/water around the ram slots good job insulating

[Yellowbeard]

Bows to the god of insulation

13 hours and no moisture rocks. That has to be some sort of record in itself.

[EnJoY]

Great clocks and timings Jody. That is no joke...6-6-5 @ over 2000MHz is just nuts... You are almost as good a memory tweaker as me now.

[3oh6]

the no swing....is because the sticks themselves are just 1-2w

too bad the imc is holding you back...

still i like the no frost/water around the ram slots good job insulating

hard to say if the IMC is holding back, could just be the sticks. i think i am going to test the 2000C7's i have here, just to see if i hit the same limits. with that said, these sticks aren't as good on air so it will be interesting to see how they translate on cold.

but yeah, on moisture after that long is so nice.

Bows to the god of insulation

13 hours and no moisture rocks. That has to be some sort of record in itself.

haha, there are no bullshyt records in this thread my friend

but thanks, overkill insulation is soooo much better than trying to trouble shoot a funky rig while benching. i'll take the OCD approach to insulation any day

Great clocks and timings Jody. That is no joke...6-6-5 @ over 2000MHz is just nuts... You are almost as good a memory tweaker as me now.

yeah, 1040MHz 6-6-5 was kind of a surprise. still want more knowing it does that at 1.85v. sooo...i reloaded:



full dewar and 20lbs'ish of dry ice for the memory. just have to decide what i want to bench tonight. look for the livestream if you are up for some boring ass endless 32M runs.

streaming the setup right now and will be benching shortly after...

http://www.livestream.com/3oh6

[skinnee]

Excellent work Jody!

Looking forward to seeing how well your plan works out with dice.

[George_o/c]

I already congratulated you like a thousand times, but I feel like it's worth it once more Congratulations

[3oh6]

haha, thanks guys. but i haven't done anything special here...just showing off Vince's nice cheap little pot. have i mentioned it is only $45USD!!!

i am streaming right now with dry ice in the pot...without acetone it is holding pot temps steady -35C ~ -40C. i have pebble sized dry ice pelletts but also crushing some into snow and it is just perfect. less work than LN2 by a mile and holds temps right where i want them. now this is what i call easy benching

boring as hell to watch, but i am gonna go grab another web cam so i can at least switch back and forth from the rig to the screen

http://www.livestream.com/3oh6

[K404]

Your SPi time at 5.05GHz is just rude

Great work Loving the insulation

Shame we cant pump so much VTT into the 980x... mine feels held-back big-time by the voltage ceiling.... or maybe im just a :banana::banana::banana::banana::banana: