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.
Click for full size...Code: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
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
Bookmarks