***ASUS Rampage IV Extreme Owners Thread***

[psyxeon]

Did you notice any vcore difference from 2105 to 3204?

[broken pixel]

----------------------------
IntelBurnTest v2.54
Created by AgentGOD
----------------------------

Processor: Intel(R) Core(TM) i7-3930K CPU @ 3.20GHz
Clock Speed: 4.66 GHz
Active Physical Cores: 12
Total System Memory: 16324 MB

Stress Level: High (2048 MB)
Testing started on 12/1/2012 3:07:31 PM
Time (s) Speed (GFlops) Result
[15:08:02] 17.961 147.1228 3.316935e-002
[15:08:32] 17.960 147.1362 3.316935e-002
[15:09:02] 18.208 145.1345 3.316935e-002
[15:09:33] 17.901 147.6173 3.316935e-002
[15:10:03] 17.863 147.9358 3.316935e-002
[15:10:32] 18.068 146.2545 3.316935e-002
[15:11:02] 18.000 146.8069 3.316935e-002
[15:11:32] 18.003 146.7809 3.316935e-002
[15:12:02] 17.938 147.3187 3.316935e-002
[15:12:32] 18.277 144.5814 3.316935e-002
Testing ended on 12/1/2012 3:12:32 PM
Test Result: Success.
----------------------------

[aussie-revhead]

Mobo is in LN2 enabled mode. Stick the pin to disable located on the top right of the mobo under the debug led.

I now have a new R4F and the same issue exists, my 3960X defaults to 1.355v and my 3930K to 1.328v, so what do you think now?

[techjesse]

Did you notice any vcore difference from 2105 to 3204?

I'm still running 2105, updating to 3204 is not viable. I'll wait for the next update and flash it if it will give astonishing results LOL

[Circaflex]

quick question, if i populate slots 1 and 4 with a gtx670 for 16x across both in SLI and if i were to use a pci x1 in the last slot for my sound card will it automatically make my second card run at 8x?

[Voodoo Hoodoo]

quick question, if i populate slots 1 and 4 with a gtx670 for 16x across both in SLI and if i were to use a pci x1 in the last slot for my sound card will it automatically make my second card run at 8x?

No, both run at x16 bandwidth. Checked that myself, using the same setup for my system.

I'm sorry I misread your question - I thought you meant using the PCI x1 slot, not putting a pci x1 in the 16 slot.

Yes, that will make slot 4 go to 8x.

Very sorry.

[broken pixel]

Got TRIM? Works!!!


http://rog.asus.com/forum/attachment...2&d=1355349903

http://rog.asus.com/forum/showthread...306#post197306

[Zaca]

Hi guys.
I need help with my overclock.
I tucked the CPU 4.8mhz, but when I do a test with prime95 or 3DMark11 or 3DmarkVantage, my PC crash in the middle of the test?s and gives a blue screen code: 0x00000101 (0x0000000000000011, 0x0000000000000000, 0xFFFFF88C03B42180, 0X000000000000000A)
I changed the following voltages:

CPU Ration: 48
X.M.P Enabled
CPU Vcore - 1.385
VTT CPU Voltage - 1.20000
2nd VTTCPU Voltage - 1.16997
CPU VCCSA - 1.20000
Dram Voltage (CHA,CHB) - 1.515
Dram Voltage (CHC,CHD) - 1.515
CPU PLL - 1.80000
PCH 1.1V . 1.10000

DIGI+Power Control:

CPU Load-line Calibration - Ultra High
CPU Current Frequency - 140%
CPU Fixed Frequency - 350
CPU Power Duty Control - Extreme

CPU VSA:

VCCSA Load-line Calibration - Extreme
VCCSA Current Capability - 140%
VCCSA Fixed Frequency - 350

When I start doing tests Prime95 seems to be okay as you can see in the picture



Please can some one help?

[Guz]

First, have you tested your mem kit with memtest? Second, imho your cpu vcore is kinda low for 4.8. Use vcore at auto with cpu LLC regular. Also no need to bump the VCSSA so much with LLC, just use regular and a fixed value such as 1.1.

[Random Murderer]

It looks like you're only loading three cores with P95 judging by your temps. Also, a few mins of P95 is not enough to verify stability. Different people will tell you different lengths of time to run it. Some say 8 hours is stable, some say 12 hours, some say 24 hours. Personally, I run it until it has made one full pass on each core, regardless of the amount of time it takes(this may not be feasible for some, as with some of my systems this has taken 36+ hours.)

That aside, I do have some tips to aid in your quest for stability(which I'm sure some may disagree with, those of you who do, speak up. I'm simply an amateur compared to a lot of you folks here on XS.)
Lower your 2nd VTT voltage, that does nothing to help with overclocking and can cause odd instabilities if raised too high. Just make sure it's within .3v of VTT and you're ok, so you should be able to set 2nd VTT to auto or whatever stock is and be fine.
Also raise your CPU current capability to 180% and frequency to 450.
Since you're running 8 sticks of RAM, up your VCCSA to around 1.225v. Also a slight bump in your PLL voltage to around 1.82v may help.
Your VCore may also be a little low, but that's on a chip-by-chip basis. You might want to just raise it by a good notch(1.42ish) for stability testing and go back and lower it later and retest for stability.

[newhit]

It looks like you're only loading three cores with P95 judging by your temps. Also, a few mins of P95 is not enough to verify stability. Different people will tell you different lengths of time to run it. Some say 8 hours is stable, some say 12 hours, some say 24 hours. Personally, I run it until it has made one full pass on each core, regardless of the amount of time it takes(this may not be feasible for some, as with some of my systems this has taken 36+ hours.)

That aside, I do have some tips to aid in your quest for stability(which I'm sure some may disagree with, those of you who do, speak up. I'm simply an amateur compared to a lot of you folks here on XS.)
Lower your 2nd VTT voltage, that does nothing to help with overclocking and can cause odd instabilities if raised too high. Just make sure it's within .3v of VTT and you're ok, so you should be able to set 2nd VTT to auto or whatever stock is and be fine.
Also raise your CPU current capability to 180% and frequency to 450.
Since you're running 8 sticks of RAM, up your VCCSA to around 1.225v. Also a slight bump in your PLL voltage to around 1.82v may help.
Your VCore may also be a little low, but that's on a chip-by-chip basis. You might want to just raise it by a good notch(1.42ish) for stability testing and go back and lower it later and retest for stability.

Sounds good to me.

I have just been perusing "ock.net" and in there, to test stability with Prime95 for just one hour is recommended as ..............

http://www.overclock.net/t/1151946/o...#post_18825344

I just tried that and get the same problem of half the workers shutting down in a random order. A suggestion in Prime95 is to change the swapfile. How do I do that? Do you think that the quoted recommendation is too severe?

I'm trying to run A 3960x C1 at 4.7 with Vcore 1.42. 16 gig of ram using the XMP setting.

I wont list all the parameters. It's safe to say that your recommendations above are about the same. I have got a 20+ LinX run with only 1.34 Vcore and over 2 hours in Primeblend.

Thanx for your time

[psyxeon]

I want to do "Per Core" Turbo overclocking, but I have no idea how to test stability.

Setting affinity for prime 95 does not work.

Any ideas?

By Per Core:

Core Ratio: 48
Core Ratio: 48
Core Ratio: 47
Core Ratio: 47
Core Ratio: 46
Core Ratio: 46


C1E: Enabled
C3 Report: Enabled
C6 Report: Enabled

Cheers

[Guz]

Sounds good to me.

I have just been perusing "ock.net" and in there, to test stability with Prime95 for just one hour is recommended as ..............

http://www.overclock.net/t/1151946/o...#post_18825344

I just tried that and get the same problem of half the workers shutting down in a random order. A suggestion in Prime95 is to change the swapfile. How do I do that? Do you think that the quoted recommendation is too severe?

I'm trying to run A 3960x C1 at 4.7 with Vcore 1.42. 16 gig of ram using the XMP setting.

I wont list all the parameters. It's safe to say that your recommendations above are about the same. I have got a 20+ LinX run with only 1.34 Vcore and over 2 hours in Primeblend.

Thanx for your time

Do you have active cooling around your mem, cpu area, specifically the boards VRMs? If not usually cores drop out due to high LLC, you can tell if this is happening if cpu speed starts to throttle while running the stress test.

[newhit]

Do you have active cooling around your mem, cpu area, specifically the boards VRMs? If not usually cores drop out due to high LLC, you can tell if this is happening if cpu speed starts to throttle while running the stress test.

Thanx for that.

No, no cooling, but I've since found out I've been using the 32 bit version of Prime . All OK with 64 bit version. I'm now getting valid use of Prime which is running past 2 hours OK (Blend).
I can now run with 4.8GHz at 1.36Vcore. I wont post all the settings but safe to say there is only CPU LLC at Ultra High taking my Vcore under load upto 1.40-1.41. I've got C1E enabled but C3,C6,C7 are disabled which might not be to everyones liking. VCCSA and VTT are at 1.2 and my 16 gig Dram is set to XMP with the MHz reduced from 2400 to 2166 for a little bit of insurance.

I will probably drop to 47x and 1.34Vcore for 24/7 use.

PS
All on aircooling...Noctua NH-D14. (I have a water rig but this setup is for a.n.other)

[Nico]

Been having what seems like alot of niggling stability issues with the RIVE. This is my third one, the other two didn't make it out of bios before having issues, and this one is an older repair replacement from Asus.

I will start off by saying when it works its fine, but if it BSoD fom to higher clk, the restart doesn't do a thing neither does the start button held down for 4secs. The only way to reset it is to turn the power supply off completely and even that can take 2-3 times before it will boot again

If I do a normal restart off windows, it will fail to boot and need to be power cycled as well, although that seems a bit better of late. Although today on restsrt it failed and started doing an irog2 update which was looking to take about 4hrs. I turned it off and when trying it a few hrs later its all good again

The only thing I have found thats a little strange was touch checking heatsink while having some of these problems, I touched the PCH sink, and noticed the SATA led lit up dimly, infact touching the board with light pressure anywhere round that corner makes the led light. Sounds like a PCB problem but I just wonder if that happens with other boards?

System details

3930k @4.8g 1.385 watercooled 60-65c
RIVE 3204 bios
64g 2133 GSKILL kit XMP settings
GTX 580
SEASONIC 1000w platinum

testing on bench

will running fine no crashes benching but tring to setup a RAM disk so I need it to be 100% reliable starting restarting etc

[Raja@ASUS]

Been having what seems like alot of niggling stability issues with the RIVE. This is my third one, the other two didn't make it out of bios before having issues, and this one is an older repair replacement from Asus.

I will start off by saying when it works its fine, but if it BSoD fom to higher clk, the restart doesn't do a thing neither does the start button held down for 4secs. The only way to reset it is to turn the power supply off completely and even that can take 2-3 times before it will boot again

If I do a normal restart off windows, it will fail to boot and need to be power cycled as well, although that seems a bit better of late. Although today on restsrt it failed and started doing an irog2 update which was looking to take about 4hrs. I turned it off and when trying it a few hrs later its all good again

The only thing I have found thats a little strange was touch checking heatsink while having some of these problems, I touched the PCH sink, and noticed the SATA led lit up dimly, infact touching the board with light pressure anywhere round that corner makes the led light. Sounds like a PCB problem but I just wonder if that happens with other boards?

System details

3930k @4.8g 1.385 watercooled 60-65c
RIVE 3204 bios
64g 2133 GSKILL kit XMP settings
GTX 580
SEASONIC 1000w platinum

testing on bench

will running fine no crashes benching but tring to setup a RAM disk so I need it to be 100% reliable starting restarting etc

The overclock you are running is not 100% stable during POST. Could be a number of things. To chase it down I'd back off on the OC, relax some memory timings. Also check your board is not shorting against your PC case. Having three boards with "issues" suggests something about your setup - either physically or you don't have a fully stable OC, which needs tuning to complete POST.


What you guys don't know is that during memory training the IMC performs over 1000 read and write operations to train the DIMMs. If the system is not fully stable with the data pattern either a channel will drop or the system will fail POST. Obviously, POST will only occur if the system is stable enough to POST, when it is, you may reach the OS and even run stress tests, but when it's not you won't. Catches a lot of people out on these newer platforms as they don't know that the memory training process is a stress test in itself. After a certain frequency some IMCs won't POST first time. Not much we can do about that. You want the system to work 100% every time your only option is to relax the overclock; memory frequency, timings and CPU frequency until the system works how you want it to.

[Nico]

Just tried it again this morning and it did BSoD while have Real Temp up and running Crystalmark, strangely enough the reset button worked, first time ever If anything I have noticed more issues while having multiple things up. I had however backed it down from 4.9ghz where it was pretty stable but would BSoD occasionally during physics tests.

Is there a good way to test mem at OC speeds these days as turbo doesn't seem to kick in with memtest 86?, mem did pass 36hrs at XMP settings,

vtt - 1.225 (really seemed to help OC at lower vcore lifting this)
vccsa - 1.10
cpu pll - 1.8 (does seem to have much effect whatever its set too)
vdimm - 1.525 (a liitle extra to keep it over 1.5)
pch - 1.125 (a liitle extra to keep it over 1.1 and to keep it in line with other voltages being up

As for the other two boards, they pretty just failed the first I left sitting in bios while I went to get a optical drive and was off and wouldn't boot when i returned. CPU died with that too.

The second was better but did the irog2 thing after it had fully complete the bios flsah and i had shut it down to connect the optical drive. should have just shut that one off too and maybe it would have come back. but tried both bios and left it running, and that too killed the second CPU.

notably neither was OC, but both where at XMP as I was going to memtest the ram next.

All tested on the bench so no case, sitting on antistaticbag on dense faom pad to give vga back panel height, test all stuff like this before installing in a case to save me later grief

It kinda points toward the ram, although it memtests fine, and I did give it a bit more vdimm and vccsa over what happly passes just for safety margin. Although still not sure why the start and restart button fail or the SATA led act up?

[Raja@ASUS]

Sounds like some kind of power short or ground loop against the heatsink of your board.

List all the components you are using (include all USB etc).

You might want to get your PSU professionally checked - I don't mean with a PSU tester. You need to find a test engineer that has an oscilloscope to probe the rails and look for any abnormalities on that side. Three boards having issues in one person's system like this points at something that you are using.

[Breit]

you killed 3 boards and 2 CPUs in a row and never thought about an error on your side? thats unbelievable...

did you know that antistatic bags are electrically conductive? i'll bet that you fried your hardware with just your ghetto-style benchtable...

[Nico]

I tested the mem at 2250 same settings for 21hrs/ 3 passes without error so seems like XMP is safe at 2133.

Sounds like some kind of power short or ground loop against the heatsink of your board.

List all the components you are using (include all USB etc).

You might want to get your PSU professionally checked - I don't mean with a PSU tester. You need to find a test engineer that has an oscilloscope to probe the rails and look for any abnormalities on that side. Three boards having issues in one person's system like this points at something that you are using.

Its extremely basic

CPU, MB, RAM and Videocard as above.

plus 300g raptor hard and an pioneer bluray (for optical)

I did consider the PSU, but haven't seen any noticeable voltage issues although as you say it could be ripple or something like that.

As for the LED thing, it happen if I press on the SATA sockets or even the corner of the board so it sounds more like a flexing issue causing a short or something, but it doesn't seem to crash it only maybe once when I first noticed it it could have cuased a restart.

you killed 3 boards and 2 CPUs in a row and never thought about an error on your side? thats unbelievable...

did you know that antistatic bags are electrically conductive? i'll bet that you fried your hardware with just your ghetto-style benchtable...

Considering I have never had boards do that before while testing exactly the same way, then I would say no, although it seem like a sequence thing where the board flakes out and is likely seeing voltages go out of control frying the CPU's. So maybe inadventantly something like the mem has caused the bios to flake out but seems a little bizzare to me.

[Nico]

Sounds like some kind of power short or ground loop against the heatsink of your board.

did you know that antistatic bags are electrically conductive?

Ok so it turns out the anti-static bag whilst good for static is not so good for power took it out from under the motherboard and the SATA led doesn't seem to act up anymore. Hopefully thats also what has been causing the random weird syptoms. Could be due to the extra weight of ram etc that causes it to be problematic when I had seen that type of thing with previous boards. Think I should probably get a proper open test bench ;p will see how it continues

Thanks for the prompts guys, sometimes it takes a few prompts to spot the issue

[XII]

People have been for far less

@Raja - RESPECT

[aussie-revhead]

Guys, I was looking at water cooling an R4E to keep me going until Ivy-E comes out but when looking at the EK R4E kit it seems like a water block is missing, is this right? The factory heatsink covers three sets of components (not including the SB) but the kit has only two blocks, what cools the components behind the I/O panel if you remove the factory heatsink? I really only want to wter cool the SB as I hate the fan noise, is there any chance a R4F heatsink will cover the MOSFETs and then I can use smaller block just on the SB? (looking at it again there isnt a block for SB only ... I dont really want to WC the whole thing coz its a mess .... thats why I got my R4F to start with)

Any help would be great.

[Random Murderer]

Guys, I was looking at water cooling an R4E to keep me going until Ivy-E comes out but when looking at the EK R4E kit it seems like a water block is missing, is this right? The factory heatsink covers three sets of components (not including the SB) but the kit has only two blocks, what cools the components behind the I/O panel if you remove the factory heatsink? I really only want to wter cool the SB as I hate the fan noise, is there any chance a R4F heatsink will cover the MOSFETs and then I can use smaller block just on the SB? (looking at it again there isnt a block for SB only ... I dont really want to WC the whole thing coz its a mess .... thats why I got my R4F to start with)

Any help would be great.

The third heatsink you speak of on the R4E doesn't actually contact anything on the board. It is there, connected by a heatpipe, strictly to help with cooling the mosfets/chipset. Same goes with the heatsink below the socket that says "Republic of Gamers" on it, it doesn't actually contact anything.
If you were to watercool the R4E, the only two blocks you would need are the mosfet and chipset blocks.

[aussie-revhead]

Awesome, thank you.