***Asus Rampage III Extreme Owners Thread***

[bassplayer]

BCLK test with Bloomfield 920 .

Cooling at Mini Cascade -80C .

http://valid.canardpc.com/show_oc.php?id=1301065

That's right . You need high PCIE in order to achieve high BCLK.
Up to 250 BCLK , 118Mhz PCIE is enough.
For 260+ BCLK , i had 126Mhz PCIE .

So that's why I can't get above 253! My PCI-E was at 115. I'll stick the 5450 in there and see if I can get higher.

Thanks man!

[skuldarin]

I tried 21 x 200; no go. Maybe 19 multi and up the BCLK? It's only my first day with an i7 - I'm a noob to this tech.

you should be able to pull off 21x205 with a voltage of 1.35 if i remember correctly. 21x210 should be doable aswell @ 1.38-1.4v.

here's my 930 results w/ 12gb of G.skill ECO CL7's at CL6 timings and OC'd to 1700mhz from 1600.

http://s3.photobucket.com/albums/y71...s.jpg&newest=1

[Utnorris]

Will the RE3 run a dual QPI 32nm Xeon?

[kpablo]

Will the RE3 run a dual QPI 32nm Xeon?

Yes it does.

[owikh84]

Now playing with R3E...

Initial test on core 4.4Ghz, uncore 4.2Ghz, dram 2100Mhz 7-7-7-20-1T



GFlops a lil bit lower than expected, need to bump up some voltages so that vCore, vtt, and vdimm can be reduced.

[KaRLiToS]

I stress test with Linx, with all my other programs running, that way, I only get 57GFlops at 4.3 Ghz (i7 930)

Should I close all of them

[OC Nub]

I stress test with Linx, with all my other programs running, that way, I only get 57GFlops at 4.3 Ghz (i7 930)

Should I close all of them

57 gflops with ht enabled is normal for 4.3. The guy with 63 gflops has ht disabled. Are you using ht?

[jcniest5]

Has the cold-boot issue been fixed yet? I remember another thread on this forum regarding that issue, I wonder if you guys are still experiencing this or not. The reason is I'm thinking of pulling the trigger to order one for myself.

[bassplayer]

Has the cold-boot issue been fixed yet? I remember another thread on this forum regarding that issue, I wonder if you guys are still experiencing this or not. The reason is I'm thinking of pulling the trigger to order one for myself.

I don't believe I'm aware of any specific cold boot issue. CBB exists for every board and is mostly reliant on the chip.

[kill_a_wat]

I have not experienced any cold boot bugs. I have only been running with H20

[Leeghoofd]

The cold boot Jcniest is referring to has got nothing to do with sub zero temps. Some boards are reluctant and will not post, only pressing resets or hard reset get's them booting... usually a bit higher QPI voltage solved it for me

[Bodar]

cbb is still there for me.. No cb though.. Lately I've been frustrated with the cpu not making proper contact with the pins.. Is there a way to make sure.. When it's not, it just haults on the memory and stays there.
Also get lots of bsod's when that happens.. Other times everything is perfect..

[KaRLiToS]

57 gflops with ht enabled is normal for 4.3. The guy with 63 gflops has ht disabled. Are you using ht?

Yes, I have HT on. Didnt know that, never tried with HT off...Im still a newb

Thank you for the answer

[bassplayer]

Has anyone benched this with a NB pot? I'll be making one in the next few days and I just took the chipset heatsink off (that yellow stuff was impossible to get off). Since it's all connected into one big heatsink, I lose all of the cooling for the VRMs and the SB if I put a NB pot on.

Any suggestions?

[tomb18]

I have a question about the 4 pin power connectors on the motherboard. Are these for increasing the current to the pci-e slots? I don't have them connected and my graphics cards seem to cause a shut down of my system when I overclock the cards.

[UrbanSmooth]

Yeah, your GPUs need power that you're not supplying, thus, you're tripping your mobo. (rough explanation)

Perhaps someone else can chime in on this and give a more technical explanation.

[skuldarin]

newer and more powerful graphics cards require more power than just what the 6 / 8 pin connections can handle sometimes. if you're using multiple cards like this all at once, you're putting a heafty load on the MB PIC-E slots... thus why the 4pin points are there. plug those in and i bet your problems will go away.

---------------------------------------------------------------------

on another note... here's my offering with the 980x on water:

oddly enough both bench marks are set to the same standard startup settings yet Vantage seems to always score higher than 06 bench for me for some reason... i dunno. at the same time, with 06, no matter what settings i put it to it ALWAYS gets roughly the same score with a difference of only around 150-250 points each time...

only thing limiting me now is my memory, but the 12gb 1866 Dominator-GT pack coming in a few days will fix that right up :P

system:
CPU - 980x @ 4.4ghz 1.38v, not 1.45v as CPUz states
Mem - G.Skill ECO's @ 12gb 1600mhz 7-8-7-24-72 2T timings w/ 1.62v
GPU - HD5970 @ 950core / 1253memory

sorry for image size, i can't seem to make it smaller without making it unreadable...

[bassplayer]

The cold boot Jcniest is referring to has got nothing to do with sub zero temps. Some boards are reluctant and will not post, only pressing resets or hard reset get's them booting... usually a bit higher QPI voltage solved it for me

Ohhh... After a bit of research, I've read (and shown to be true) that this only happens to people who used Gigabyte boards before the RIIIE. Gigabyte won't admit it, but their first X58 boards applied VTTPWRGood that was too high. Because of this, the chips degrade and will only work in a board if the VTT is high enough.

[Utnorris]

So for whatever reason I cannot complete the second CPU test in Vantage, the one with the planes flying into the suspended circles. It's Linx stable, runs BOINC all day with no issues and I can play games all day, but if I try and run a Vantage benchmark it hangs there. I can hit the escape button and it stops the test, but for whatever reason it just hangs there and stays on the "warming up" part. I can even see my cpu throttling, but still no go. Any ideas?

[donmarkoni]

Yeah, your GPUs need power that you're not supplying, thus, you're tripping your mobo. (rough explanation)

newer and more powerful graphics cards require more power than just what the 6 / 8 pin connections can handle sometimes. if you're using multiple cards like this all at once, you're putting a heafty load on the MB PIC-E slots... thus why the 4pin points are there. plug those in and i bet your problems will go away.

6 / 8 pin connectors can supply enough, problem is the MB. Thin lines on the PCB that cannot supply enough through PCI-E if needed. That is why extra power connectors are close to PCI-E, so the power can be supplied via thick short lines.

Ohhh... After a bit of research, I've read (and shown to be true) that this only happens to people who used Gigabyte boards before the RIIIE. Gigabyte won't admit it, but their first X58 boards applied VTTPWRGood that was too high. Because of this, the chips degrade and will only work in a board if the VTT is high enough.

Are you referring that Gigabyte MB's supply to much voltage on VTT or VTTPWRGOOD? Don't mix those two.
If you are talking about VTT, it is known that using XMP on any board and especially on GB, it will set extreme VTT (1.55-1.65V). That's why people degrade their CPU's.
If you are talking about VTTPWRGOOD, which is just a signal line and not a power supply line, then that is a major issue.

[Leeghoofd]

Has anyone benched this with a NB pot? I'll be making one in the next few days and I just took the chipset heatsink off (that yellow stuff was impossible to get off). Since it's all connected into one big heatsink, I lose all of the cooling for the VRMs and the SB if I put a NB pot on.

Any suggestions?

Don't waste ya time with subzero, watercooling or a good air cooler are more than enough. IMC is in the CPU...

So for whatever reason I cannot complete the second CPU test in Vantage, the one with the planes flying into the suspended circles. It's Linx stable, runs BOINC all day with no issues and I can play games all day, but if I try and run a Vantage benchmark it hangs there. I can hit the escape button and it stops the test, but for whatever reason it just hangs there and stays on the "warming up" part. I can even see my cpu throttling, but still no go. Any ideas?

It's not stable for Vantage... Lower your clocks, add more cold or pump volts

[Raja@ASUS]

If you are talking about VTTPWRGOOD, which is just a signal line and not a power supply line, then that is a major issue.

He is talking about VTTPWRGOOD pull-up deterioration to the extent where the voltage needs to be disproportionately high for the processor to POST. The reason I think it was being done was to encourage the early i7 processors to POST in the said boards regardless of the applied VTT; this had an unfortunate side-effect over time as some users unwittingly found out. BTW, this was checked in a HW lab, so is not solely based on forum level conjecture.

[Visard]

So for whatever reason I cannot complete the second CPU test in Vantage, the one with the planes flying into the suspended circles. It's Linx stable, runs BOINC all day with no issues and I can play games all day, but if I try and run a Vantage benchmark it hangs there. I can hit the escape button and it stops the test, but for whatever reason it just hangs there and stays on the "warming up" part. I can even see my cpu throttling, but still no go. Any ideas?

Had the same problem with 980X - Just update with the Vantage Patch 1.0.2 !

[Toritechie]

6 / 8 pin connectors can supply enough, problem is the MB. Thin lines on the PCB that cannot supply enough through PCI-E if needed. That is why extra power connectors are close to PCI-E, so the power can be supplied via thick short lines.

I'm running 2 GTX295's, and I have been running with the 4pin near the CPU connected (EZPLUG 1). I've also tried running the one down low (EZPLUG 2), but if I do that, my system will not boot. It's a fiddly pin to put in, on my Cosmos 1000 it sits right near the PSU and is a pain in the butt. I'm assuming I'll be OK with just one of the 4-pins going?

[facboy]

Ohhh... After a bit of research, I've read (and shown to be true) that this only happens to people who used Gigabyte boards before the RIIIE. Gigabyte won't admit it, but their first X58 boards applied VTTPWRGood that was too high. Because of this, the chips degrade and will only work in a board if the VTT is high enough.

mine posts, sort of, but keeps resetting constantly. it was working fine and then rebooted spontaneously and never posted successfully afterwards. it just keeps looping, gets to about 1 second after successfully posting (usually the CMOS error screen), and then resets. brand-new 980x on a brand-new R3E, has been RMA'd once after failing after 6 weeks, RMA returned with new mobo (not sure about cpu) and worked for 6 hours and failed in exactly the same way - rebooted, then reset cycling again.