SuperPi32m 5Ghz Ivy All Out Challenge!

[Raja@ASUS]

I did look up the Intel datasheet amongst other things the last time Raja mentioned tRRD/tFAW 4/16 earlier in this thread & although it could be a BIOS/Software report issue its less likely as the apparent performance increase has been found on both my Gigabyte & M.Beier's Asrock so I'm inclined to say there may be more to it.

Well unless someone is breaking DDR3 timing rules I cannot see how that is possible. It is a 64 bit wide channel Chances are there is disparity between actual value and displayed value.

I agree though, whatever works is okay, just does not hurt to throw some understanding at things sometimes

[Splave]

^ exactly, when I started it was very overwhelming to trying and lower everything but now starting to understand the meaning makes things easier

[Splave]

3 more chips coming today, hopefully can bench over 2600 atleast one

also upgrade ripjaws bbse skin, epoxied to murdered out dominator skin will get to freeze tomorrow if I am lucky

[trans am]

All the good stuff is here thanks to Stasio http://forums.tweaktown.com/gigabyte...ity-tools.html

thanks buddy! thanks stasio.

[C-N]

upgrade ripjaws bbse skin, epoxied to murdered out dominator skin

I was thinking of doing something like this to try to cobble my single stage on them as Raja mentioned a few pages back. Guess the spreaders are on there for life with epoxy was the epoxy anything special?

thanks buddy! thanks stasio.

You're welcome

I re-tested tRRD 4 & tFAW 16 & all the way down to 3 / 10 & can find no gain or loss at all @ 8.11.7 2550 Cant explain why I appeared to get a boost when I last tested all be it at 9.11.8 2450..... Guess that blows that one for me lol

[Splave]

^negative, nothing special. and its easy to remove the spreaders just torch them for like 2 seconds and they pop right off.

[Raja@ASUS]

Yeah, just be careful when pulling them off. Would recommend you guys remove the epoxy on the heatspreaders as that stuff is not very thermally conductive. Corsair rely on the PCB copper plane at the top of the modules for thermal transfer rather than the epoxy.

[Martin.v.r]

win7 64 bit ULtra no tweak.
I7 3770k 5GHZ 1.4v watercooling
ASUS Maximus P67 B3
g.skill ripjaws-x 17600cl7-10-10-27 2T PSC 2x2GB
8-10-8-27 1T 1.65v 2200MHZ air cooling
32m 06m 38.440s

[M.Beier]

Yeah, just be careful when pulling them off. Would recommend you guys remove the epoxy on the heatspreaders as that stuff is not very thermally conductive. Corsair rely on the PCB copper plane at the top of the modules for thermal transfer rather than the epoxy.

Hmm, are you suggesting a gain by removing the heatspreaders, and remounting with lets say.... ceramique or MX2-3-4 ? I have honestly not considered removing the spreaders from any Dom.... But, if there is a gain, then it might be worth doing it.

[C-N]

The way i read it is epoxy dont transfer heat too good but imo that dont mattermuch as splave will be chilling them & be careful not to rip the ic's off the pcb ...

[M.Beier]

The way i read it is epoxy dont transfer heat too good but imo that dont mattermuch as splave will be chilling them & be careful not to rip the ic's off the pcb ...

Ripping off IC's is the main reason why I have not done it.... I know it is possible, but there is always a risk...

[SoF]

Neeeeed more time with MVF...but at least sub thiiiiiiiirty and hunting tweeeeeeeeeeeeeeeeeeenty now

[dinos22]

looking badass C-N

[littleBoy]

windows xp 32bit sp3 little-tweak.
I7 3770k 5GHZ<104.2 x 48> 1.416v watercooling
ASRock z77 Extreme6
g.skill ripjaws-x 17600 cl7-10-10-27 2T PSC 2x2GB
8-10-9-27 1T 1.74v 2200MHZ air cooling
32m 06m 23.672s

[Raja@ASUS]

Hmm, are you suggesting a gain by removing the heatspreaders, and remounting with lets say.... ceramique or MX2-3-4 ? I have honestly not considered removing the spreaders from any Dom.... But, if there is a gain, then it might be worth doing it.

No, don't replace the epoxy if you are going to be using the Corsair Dominator PCBs with those heatspreaders as the PCB copper plane acts as a good enough transfer interconnect. If you intend to use those heatspreaders with any other vendor module then removing the epoxy residue from the heatspreaders is worthwhile. It's not very thermally conductive and I found with my cascade that the PSC modules I used would not clock to DDR3-2600, as the transfer to the ICs was very poor, with the epoxy basically acting as an insulator. Cleaned off epoxy (epoxy with TIM did not work well) and used good TIM and my BBSEs cold bugged with the PSC ended up hitting DDR3-2600 at CAS 6. It will chip off with a screwdriver and then you can replace it with a decent enough TIM. Hopefully this will take all the guesswork and armchair supposition out for you guys.

-Raja

[M.Beier]

No, don't replace the epoxy if you are going to be using the Corsair Dominator PCBs with those heatspreaders as the PCB copper plane acts as a good enough transfer interconnect. If you intend to use those heatspreaders with any other vendor module then removing the epoxy residue from the heatspreaders is worthwhile. It not very thermally conductive and I found with my cascade that the PSC modules I used would not clock to DDR3-2600, as the transfer to the ICs was very poor with the epoxy basically acting as an insulator. Cleaned off epoxy (epoxy with TIM did not work well) and used good TIM and my BBSEs cold bugged with the PSC ended up hitting DDR3-2600 at CAS 6. It will chip off with a screwdriver and then you can replace it with a decent enough TIM. Hopefully this will take all the guesswork and armchair supposition out for you guys.

-Raja

Ahh okay, super, well, was worried that I lost some cooling when plugging on my LN2 rampot, which eventually I'll end up doing in this challenge it seems.... I wont stand a chance without bringing them -150C

[SoF]

Hmmm lower baseclock is better - I am surprised ^^

[Vivi]

Max IMC on air is 2760, so ill try for more tonight with CPU on ln2

Vivi - 6:20:656 - 49x102 - 1360 - 9-12-12-22-110-1 - Gigabyte UD5H - Team Xtreem LV 2600 CL10 - 1.88v - stock air

[Splave]

noice

[C-N]

looking badass C-N

thanks man.... the board is ace

Max IMC on air is 2760, so ill try for more tonight with CPU on ln2

Vivi - 6:20:656 - 49x102 - 1360 - 9-12-12-22-110-1 - Gigabyte UD5H - Team Xtreem LV 2600 CL10 - 1.88v - stock air

Looking great with the LV2600's, I gave mine a quick spin but nothing like as good as that.

### Edit ###

C-N - 6:19:719 - 47x106.4 - 1276MHz - 8-11-7-22-70-1 - Gigabyte UD5H - GSkill C8 2133mhz 2x2GB/BBSE - 1.91v - Stock Air

[dinos22]

man you got the bandwidth dialied in on that 3570, im guessing the lack of cache is causing latency to drop off. magic mate!





GOOOOOOO VIVI

[OC Nub]

Nice scores C-N and Vivi!

[Rosty]

Vivi and C-N nice work

[C-N]

Previous post updated with a tiny update as a result of my now mandatory quick few runs in the morning before going to work!!!! lol

man you got the bandwidth dialied in on that 3570, im guessing the lack of cache is causing latency to drop off. magic mate!

Yup she's running strong. Thats all I can think the lower 3570K cache must be effecting the latency. I can get it down way lower just cant make it stable lol

Nice scores C-N and Vivi!

Vivi and C-N nice work

Thanks guys

[M.Beier]

Honda forgot about this thread?