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

[Xel_NaGa]

Even though my BIOS only sees 2 sticks?

Surely, with a brand new 980x installed today; it can't be a CPU issue?

I'm tempted to just pick up a different brand of RAM and sell on these Tridents. They work flawlessly with the 930.

I'll try some different settings, but I've already set a variety of QPI's and other RAM setting,s and am now using BIOS 1207. All to no avail.

Would 0003 be any different?

I'm assuming this issue is happening at stock speeds?

if so something to try here... I noticed that this motherboard for no reason will get wonky with advance ram timings. Try dropping in your other cpu copy down all the timings with your basic x-x-x-x xt manually coded. then use those timing with your 980 and see if that helps. My 980 does wonky stuff like that during overclocking with different uclk speeds.

another thing to try... are you running 1.5x memory spd uclk? or 2x memory spd uclk? if your running 2x back it down to 1.5x see if that helps.

[Toritechie]

I'm assuming this issue is happening at stock speeds?

if so something to try here... I noticed that this motherboard for no reason will get wonky with advance ram timings. Try dropping in your other cpu copy down all the timings with your basic x-x-x-x xt manually coded. then use those timing with your 980 and see if that helps. My 980 does wonky stuff like that during overclocking with different uclk speeds.

another thing to try... are you running 1.5x memory spd uclk? or 2x memory spd uclk? if your running 2x back it down to 1.5x see if that helps.

Yep, this is happening at stock speeds. I have also tried both 1.5x and 2x uclk; no luck.

Channel A - the first DIMM I use, cannot be discovered and I cannot even enter BIOS when I have a single stick in this channel. I've tried all three sticks in this channel, nothing.

I've also tried some other DDR3 from another machine in Channel A. Only B and C are recognizable.

However, A is discoverable and works fine when I use my old CPU.

What's the chance that the 980x Intel sent me for an RMA is damaged? I guess the only way to really work this out is to drop the 980x into another 1366 board and test; unfortunately I don't know anyone else around here (rural area) with a 1366 board.

I can't think of anything else I should do - the 980x just does not want to run channel A. I've reseated it twice, and can't discern any problems with the pins etc.

Could the R3E be the problem? But if so, why isn't it the problem when using a 930? Gah.

[zsamz_]

you lost channel A on the cpu
a had a 920 that lost channel B
no matter what ram i put on that channel ....
you gonna have to rma the cpu

Yep, this is happening at stock speeds. I have also tried both 1.5x and 2x uclk; no luck.

Channel A - the first DIMM I use, cannot be discovered and I cannot even enter BIOS when I have a single stick in this channel. I've tried all three sticks in this channel, nothing.

I've also tried some other DDR3 from another machine in Channel A. Only B and C are recognizable.

However, A is discoverable and works fine when I use my old CPU.

What's the chance that the 980x Intel sent me for an RMA is damaged? I guess the only way to really work this out is to drop the 980x into another 1366 board and test; unfortunately I don't know anyone else around here (rural area) with a 1366 board.

I can't think of anything else I should do - the 980x just does not want to run channel A. I've reseated it twice, and can't discern any problems with the pins etc.

Could the R3E be the problem? But if so, why isn't it the problem when using a 930? Gah.

[Toritechie]

you lost channel A on the cpu
a had a 920 that lost channel B
no matter what ram i put on that channel ....
you gonna have to rma the cpu

Well ...

For a 4th time, I reseated the 980x ...

and this time ...

I can detect 6GB. All good. Triple channel.

It must have been a pin contact being slightly off.

Woooooo!

[skuldarin]

i still can't seem to get past the 4.5 mark on my 980x...

just curious though... does anyone seem to have issues with clocks between 4.5 and 4.599 but yet do fine with 4.6+?

i ask because i can run 4.4 perfectly stable it seems and barely scratch at 80*c for full load temps with about 1.40625v and about 1.3125v qpi. as soon as i go to 4.500.5 (this being the bare minimal mark) vCore wants to be like 1.46-1.47v and temps sky rocket WAY out of ratio for only 100mhz...

EDIT: also, it seems if my QPI voltage goes any higher than about 1.33, my system refuses to save that setting on startup and doesn't show a display at all... just black screen (no signal). funny part is, i thought that maybe it was due to IOH needing voltage or increased pcie to like 103-105 range... but it's not an issue of posting. it still starts up and goes to windows without the displace like it's not intializing video bios...

[Maxpayne]

hello, I have one question
I have also Asus R3E, I want to buy a sound card X-fi Titanium HD and graphics card Asus 570gtx DirectCU II
The question is: Will the video card very warm sound card?
Video card will have a very big temperature they do not have good space for exhausting air?

sorry for my English

[donmarkoni]

So many posts in one day... A lot to catch up.

Does 0003 have this feature? I haven't gotten the opportunity to flash to it yet.

Yes, 0003 doesn't have Sync Mode any more, so it's probably same as 1207.

I think 0003 is 1207 but with tRL option added for each channel. tRL being too tight (or set wrong) can cause channels to be dropped.

And I think that your thinking is correct.

Well ...
For a 4th time, I reseated the 980x ...
and this time ...
I can detect 6GB. All good. Triple channel.
It must have been a pin contact being slightly off.
Woooooo!

I'm sorry for not helping as you addressed to me, I was very busy, but I'm glad you sorted it out and that there are many smart people here willing to help right people with right attitude.

[tistou77]

with memory Elpida Hyper, it's worth testing the 0003 (if same as 1207 with the option tRL)?

thanks

[Mydog]

Hello guys I'm scratching my head here, I've just switched from the R2E to R3E and I've got a little problem. I had one 5970 and two 5870 running on my R2E but after the switch CCC only detects the 5970 in the first pcie-slot but sees one of the 5870's and says its not propurly connected with a cfx bridge, device manager also only sees one of the 5870's. Now for the tricky part they are all watercooled and not so easy to swap around and I suspect that either pcie-slot2 are dead or the card is dead and the 5870 in slot 3 is alive but not connected directly to the 5970 so cfx is not working.
How do I test this?
I've connected both of the extra power connectors and swapped the pcie powercables around.

Pic of the setup

[Toritechie]

I'm sorry for not helping as you addressed to me, I was very busy, but I'm glad you sorted it out and that there are many smart people here willing to help right people with right attitude.

Hey Don, it's ok. I'm surprised it was a seating issue, it's not like I don't know how to align a processor in the socket. I think the tension was too high on my waterblock - and I've read that it can have a negative affect on a CPU being able to run all RAM channels. I suppose the tension screws with some pin alignments?

Right now I'm trying a few variations of templates, and am looking at getting 4.6 stable with the 980x. My Trident RAM doesn't seem to like 2000mhz; so I'm running in the range of 1600-1800 with multipliers of 23 - 27, and an uncore ratio of 1.5x.

I've got 4.45 rock solid, still fiddling around to get the 4.5 - 4.6 good. But, it's my first day with the 980x.

I have to thank Intel also. I bought a second-hand 980 CPU from Ebay, only to find it dead on arrival. I contacted Intel here, and did an RMA. Total turnaround for the RMA was about 1 week.

I later discover that the seller had done a cold run (LN2) for 30 minutes .... on the CPU I bought from him .... lol.

Anyway; I don't have any hard feelings against the seller, he was helpful and willing to refund me.

Just good to see the professionalism of Intel, and that they are willing to support their high end products.

[skuldarin]

Toritechie - if you get 4.5-4.6 going decently, could you post your settings for it? i can't seem to get my 980 to go over 4.4 for the life of me. i've tried countless BIOS and a good number of different bclk's and multipliers and hours of voltage play. no matter what i do, i can't seem to get 4.5+ without BSOD'ing.

[CryptiK]

Skuldarin if your CPU needs ~1.31v for 4GHz stable your chances of 4.5GHz stable at sane vcore are slim to none IMO. Max VID on 32nm 980X is 1.375v and abs max is 1.40v so be careful when pushing for 4.5.

[PC400]

Oh god, read please !

http://www.xtremesystems.org/forums/...d.php?t=268385

Help-me

[skuldarin]

Skuldarin if your CPU needs ~1.31v for 4GHz stable your chances of 4.5GHz stable at sane vcore are slim to none IMO. Max VID on 32nm 980X is 1.375v and abs max is 1.40v so be careful when pushing for 4.5.

actually, 1.3125 vCore was just to get 4ghz to pass in LinX, it runs perfectly stable at 1.28 rather easily.

4.4ghz takes about 1.38-1.39 vCore and my QPI stays the same between 4 and 4.4 @ 1.3125v. it isn't until 4.5+ that my QPI needs any increase.

[p3gaz_001]

guys, do the new bios has new features on the fan control?! i've discovered that... the cpu fan, either i set it on manual silent or turbo duty mode, or what ever the cpu fan speed is always at 2000rpm, (the max fan speed) should work like that, other fans does not behave like that, all fans are the same, cm 2000rpm led blue / red, so what's wrong?! is fan connected on the cpu failing, or what?! ... this is just one of the strange behavior i've been whatching from 2 weeks till now!! need some help ...

[spoolindsm127]

Yes, 0003 doesn't have Sync Mode any more, so it's probably same as 1207.

Thanks will have to give 0003 a try after 1207. I also can't get past 4.46GHz

guys, do the new bios has new features on the fan control?! i've discovered that... the cpu fan, either i set it on manual silent or turbo duty mode, or what ever the cpu fan speed is always at 2000rpm, (the max fan speed) should work like that, other fans does not behave like that, all fans are the same, cm 2000rpm led blue / red, so what's wrong?! is fan connected on the cpu failing, or what?! ... this is just one of the strange behavior i've been whatching from 2 weeks till now!! need some help ...

CPU Fan Header is only header on the motherboard that requires you to use a PWM (4-pin) fan for Q-Fan speed control. If the fan plugged into CPU Fan Header is not PWM it will always run full speed.

Hope that helps

[CryptiK]

actually, 1.3125 vCore was just to get 4ghz to pass in LinX, it runs perfectly stable at 1.28 rather easily.

4.4ghz takes about 1.38-1.39 vCore and my QPI stays the same between 4 and 4.4 @ 1.3125v. it isn't until 4.5+ that my QPI needs any increase.

4GHz requires 1.28v and 4.4GHz takes 1.39v say...interesting vcore scaling. That's 0.0275v per 100MHz which is less than usual from my experience. Sure thats comparing fully stable 4GHz to fully stable 4.4GHz?

Might be board, we know these things can be picky for some people, but more than likely its the CPU just reaching the non-linear scaling point.

[skuldarin]

4GHz requires 1.28v and 4.4GHz takes 1.39v say...interesting vcore scaling. That's 0.0275v per 100MHz which is less than usual from my experience. Sure thats comparing fully stable 4GHz to fully stable 4.4GHz?

Might be board, we know these things can be picky for some people, but more than likely its the CPU just reaching the non-linear scaling point.

ya fully stable, 4ghz and 4.4ghz running stable at 50 passes in LinX with full memory at 1866 w/ CL7 timings (7-9-7-24-72-2T / 58-59-60 RTL) with 6 sticks at 12gb. i've tried going to 1T but the voltages required for that is far beyond what i'd allow (i've gone to 1.82v and still doesn't see more than 10gb, that includes increasing CL and RTL times to be obscenely slopy).

i've been thinking it's the peak of the scale point also because no matter what bios it is, no matter what base clock VS. multiplier it is, it WILL NOT go to 4.5 or above for whatever reason. so i'm going to settle with 4.4 being the peak, i rolled it back down to 4.25ghz for daily use... of which those speeds are pretty respectable for daily use (4.4 is daily capable, but with the coming summer months 4.25 is a much lower temp hazard as ambient temps are normally around 85-100 and high humidity).

i will test my sister tower (same exact build as this one that i built for my father) and see if it's 980 is any better. if it is... i might just do an old switch-a-rooo as the old man will never need more than 4ghz for his setup.

[p3gaz_001]

CPU Fan Header is only header on the motherboard that requires you to use a PWM (4-pin) fan for Q-Fan speed control. If the fan plugged into CPU Fan Header is not PWM it will always run full speed.

Hope that helps

gosh the cpu fan is a normal 3 pins fan ..... so must buy a pwm fan

thanks a lot

[CryptiK]

Arctic Cooling make a nice 4 pin PWM fan, used to use one back when I was on aircooling.

Edit: This one

You want the F12 (120mmx25mm) model.

[Xel_NaGa]

Well ...

For a 4th time, I reseated the 980x ...

and this time ...

I can detect 6GB. All good. Triple channel.

It must have been a pin contact being slightly off.

Woooooo!

good thing it was an easy fix

[CryptiK]

Well I fixed my cold boot issue...........................with a new CPU Also fixed the fact I only had 4 cores...........

Just testing so far but looks good.

Prime95 1hour | 4GHz @ 1.171v

[zoson]

I'm stuck at 200bclk.
I can never fully stabilize at anything above.
Current settings:
4210MHz
Mult:21
BCLK: 200
PCI: 100
CPU Config: all disabled, except HT and ALL cores enabled.
vCore: 1.3750
vTT: 1.33750
CPU PLL: 1.2
vIOH: 1.20v
vICH: 1.16v
vDDR: 1.65625v
DDR: 2000MHz 3x4GB 9-9-9-25-110-1T

With the above settings I'm fully 100% LinX stable with both ALL mem and 8GB mem. Running 8GB results in 10C higher temps during LinX runs than ALL mem does.

If I even go up one mhz to 201BCLK, it seems no matter what I do, I can not get stability. Linx will stop on error after the 2nd loop almost every time.
*sigh*

I was able to boot at 25x180 with 1207 bios tho. That was not possible with any bios before for me.

[Fire4Wire]

guys listen

got beta bios 1301 today with following fixes


Updated SB RAID Option ROM to v10.1.0.1008

stay tuned for next official bios update with these fixes

[Exupfizzer]

can you share the new bios 1301 please

guys listen


got beta bios 1301 today with following fixes


Updated SB RAID Option ROM to v10.1.0.1008

stay tuned for next official bios update with these fixes