DFI UT X58-T3EH8 continued - Part 2

[JustCallMeFrank]

What happens when you unlock the multiplier on a w3520? You get into Windows >@6GHZ
w3520 @6GHZ (215x28)


LOL, it was an attempt on @4.3GHZ(215x21), but somehow it ended up with x28. It was fun as long as it last.

How did this happen...? I have a W3520 too..... please... do telll....

[p2501]

This didn't happen. That's a CPUZ error, nothing else. Still, looks good.

Well, the saga ended unhappily. Was net browsing yesterday when my rig locked up, and seconds later the smell of burning electronics filled the room. Yanked the power cord, opened the case, streams of grey smoke filling the case. Pulled everything out, unplugged everything, attached to another PSU and tried booting with one stick of RAM and old S3 Virge PCI backup card, streams of grey smoke poured from the cap near the EPS 8-pin connector and the board displayed random garbage on the LED debugger (not whole letters or numbers, just fragments lighting up).

Now to see how long warranty takes, seeing as the board was bought from Newegg in the USA earlier this year, and I'm now back in Australia again.

Man good luck with RMA, you really are struck with a run of bad luck aren't you?! Maybe this says something about the boards quality. Only good thing is that it was stable for a while..

[eva2000]

Well, the saga ended unhappily. Was net browsing yesterday when my rig locked up, and seconds later the smell of burning electronics filled the room. Yanked the power cord, opened the case, streams of grey smoke filling the case. Pulled everything out, unplugged everything, attached to another PSU and tried booting with one stick of RAM and old S3 Virge PCI backup card, streams of grey smoke poured from the cap near the EPS 8-pin connector and the board displayed random garbage on the LED debugger (not whole letters or numbers, just fragments lighting up).

Now to see how long warranty takes, seeing as the board was bought from Newegg in the USA earlier this year, and I'm now back in Australia again.

ouch sorry to hear mate.. take photos and all of the damage - good to document everything before sending back for RMA

Is that Corsair Dominator GT 8-8-8-24 2000Mhz kit one of the dodgy ones? I just bought that kit about a week and a half ago. I'm still waiting for other parts to arrive though so I haven't been able to try it out yet though.

I am going to use it with a DFI T3EH8 and a W3520. So what happens if the RAM stuffs up?

Corsair will take care of you if you need to RMA, 3 options for RMA listed at http://i4memory.com/f74/corsair-domi...ncement-18738/

ORIGINAL: Corsair_Mike

For Dominator GT/Elpida Hyper IC customers with failed parts we offer the following 3 options.

1. A refund, original invoice required

2. A set of hand screened Dom GTs (Elpida Hyper ICs), with full warranty of course

3. Replacement with a 12GB 1600 kit: http://www.corsair.com/_datasheets/HX3X12G1600C9.pdf

[SoulsCollective]

what pwm temp you had under a load?
and what overclock?

PWM temp I'm really not sure about exactly, never measured it, but the cooler over that area wasn't all that hot - not cool, there was heat transfer, but not burning up either. Overclock was never more than 3.7 - see above, I could never get more than 185 BCLK to boot.

[p2501]

Just making sure here, for VTT/IMC/Mem stability checks I need to run Prime with FFTs ranging from 128 to 4096 with in-place unchecked and max memory right? Or whats the best setting for this purpose?

[ci2kla]

Just making sure here, for VTT/IMC/Mem stability checks I need to run Prime with FFTs ranging from 128 to 4096 with in-place unchecked and max memory right? Or whats the best setting for this purpose?

Just blend it is ok

[p2501]

Blend with mem maxed right?

[Razer(x)]

Blend with mem maxed right?

As you want, blend with default settings is ok too, but i always prefer to give it all the memory

[p2501]

Thanks guys. Btw, Razer I think I'm on to something. Running 12GB was as easy as slapping the additional 6GB in and just run them. I think most of my problems came from the fact that asking for LinX stability at 89°C was just a tad too much. Even in a warm room (sry have no thermometre around) I'm not going over 70°C now at 1.32 vcore. You just have to love water cooling, 50 LinX passes max mem easy even at 7-7-7-24-1N. I'm posting some screenies when I confirmed prime stability.

TRUE has met it's end for people that have no air conditioning around in the summer for benching or stability testing me thinks.

[Razer(x)]

Thanks guys. Btw, Razer I think I'm on to something. Running 12GB was as easy as slapping the additional 6GB in and just run them. I think most of my problems came from the fact that asking for LinX stability at 89°C was just a tad too much. Even in a warm room (sry have no thermometre around) I'm not going over 70°C now at 1.32 vcore. You just have to love water cooling, 50 LinX passes max mem easy even at 7-7-7-24-1N. I'm posting some screenies when I confirmed prime stability.

TRUE has met it's end for people that have no air conditioning around in the summer for benching or stability testing me thinks.

Uhm, mine isn't a temp problem, maybe PWM, but surely not cpu. I've noticed that if i load a saved profile i'm unstable 90% of times, no problem with setting it manually, bah!

[p2501]

Uhm, mine isn't a temp problem, maybe PWM, but surely not cpu. I've noticed that if i load a saved profile i'm unstable 90% of times, no problem with setting it manually, bah!

Did you play with boot-up base clock or are you running that on auto?

[Razer(x)]

Did you play with boot-up base clock or are you running that on auto?

Auto.

[p2501]

Auto.

Try around with that setting, for me it also had an impact on stability as I discussed that with Eva.. somewhere in this thread I think. Also PWM frequency setting and IOH volts. When you set a boot-up baseclock there's no jumping RTL anymore I think. But are your issues now gone with your new ram or not?

Edit: Found it. But maybe you've read that already.

[Razer(x)]

Try around with that setting, for me it also had an impact on stability as I discussed that with Eva.. somewhere in this thread I think. Also PWM frequency setting and IOH volts. When you set a boot-up baseclock there's no jumping RTL anymore I think. But are your issues now gone with your new ram or not?

Edit: Found it. But maybe you've read that already.

In part they did. So, i'm at CAS8, gonna try 190-195

[hel_19]

the best bios?

[Razer(x)]

the best bios?

The latest

[hel_19]

The latest

my experience Xeon W3520

Stability
bios 4/28 205x19 1.18v LinX, prime95
bios 6/19 205x19 1.18v not stable

mult x20 bad, 195x20 1.25v not stable

Thank you

[ci2kla]

my experience Xeon W3520

Stability
bios 4/28 205x19 1.18v LinX, prime95
bios 6/19 205x19 1.18v not stable

mult x20 bad, 195x20 1.25v not stable

Thank you

What?! 0619 BIOS works flawlessly with my w3520...1.175v BIOS to pass 200x20 no problem. However with high uncore (4000Mhz) will need slightly higher vcore to compensate. 227 bclk with QPI enabled cold boot to windows without any problem, but 228 failed no matter what PCI-E I used I guess that's the bclk wall for my w3520 then, can't handle high QPI Frequency. With QPI Disabled 232 bclk cold boot is ok.

[p2501]

Well, here it is. Not really amazing voltages, but stable.


vcore load 1.32
IOH/ICH 1.19
RAM 1.65
VTT load 1.43
Boot-up bclk 185
DRAM PWM reduce 15%
(only listing changes, also 7-7-7-24-1N needed to be put in manually)

[Razer(x)]

Wow! Really nice temps!

PS: My problem seems to be still here

[StimulatedBoy]

my experience Xeon W3520

Stability
bios 4/28 205x19 1.18v LinX, prime95
bios 6/19 205x19 1.18v not stable

mult x20 bad, 195x20 1.25v not stable

Thank you

BIOS 04/28 and 05/08 working flawlessly on my system with Xeon W3570 whereas 06/16 and 06/19 is a no go OC-wise...

[rioja]

Well, here it is. Not really amazing voltages, but stable.


vcore load 1.32
IOH/ICH 1.19
RAM 1.65
VTT load 1.43
Boot-up bclk 185
DRAM PWM reduce 15%
(only listing changes, also 7-7-7-24-1N needed to be put in manually)

your temps are very low actually
my xeon 4300 1.36v run with temps 76C under prime and 79C under linpack
and this is watercooled by heatkiller and pa120.3 with liquid metall as tim!
i would like oc more but i have to stack at 205x21 with current cpu

[p2501]

your temps are very low actually
my xeon 4300 1.36v run with temps 76C under prime and 79C under linpack
and this is watercooled by heatkiller and pa120.3 with liquid metall as tim!
i would like oc more but i have to stack at 205x21 with current cpu

Thanks!
My loop is like DDC3.2 xspc res top -> 2 PA120.3 with medium yates -> HK3.0 Cu -> mcw30, all with 3/4" Norprene. Apart from the temps this build was just so fun, but I already have things I'd do differently with the next build.
If I have the time I would want to post some pics in the watercooling build thread.

[Razer(x)]

Thanks!
My loop is like DDC3.2 xspc res top -> 2 PA120.3 with medium yates -> HK3.0 Cu -> mcw30, all with 3/4" Norprene. Apart from the temps this build was just so fun, but I already have things I'd do differently with the next build.
If I have the time I would want to post some pics in the watercooling build thread.

Nice, i wanna add a second radiator to my PC too.

I'm waiting 4 the photos

[jagt]

BIOS 04/28 and 05/08 working flawlessly on my system with Xeon W3570 whereas 06/16 and 06/19 is a no go OC-wise...

What do you mean by "no go"? Does 16 & 19 require more voltage than for example 05/08? Just got a new board from RMA so I'm still testing. So far I've established that my previous stable vcore setting is not enough for 06/19. I've flashed back to 05/08 now to check if it improves.