Sorry Eva, I meant 0507.
Sorry Eva, I meant 0507.
asus said the xeon 55xx series aren't meant for oc'ing so limited support. But Xeon W3520/3540/3570 have full support from 0007 and 0001 bioses. Throttling haven't noticed but then i've mainly tested on non-turbo multipliers so far for W3540. But 0001 bios fixes turbo throttling apparently but 0007 doesn't which is bios i'm still on.
same answer applies - try 0601 and 0007 bios
---
I'm not liking 0507 and I'll take your advice and move to 0601. I think I figured out how to make cpu tweaker work though:
1T: http://i188.photobucket.com/albums/z...weakers31T.jpg
2T: http://i188.photobucket.com/albums/z...weakers32T.jpg
Ignore the 4tCL, its not reading right and is 9. I also have C-States enabled.
If they are changing, Everest doesn't show the right timings but cpu-z does. All I did was put the system into S3 sleep and upon waking up I was able to change the timings. I'm not sure what to expect from a timing change from the bench in Everest. Does that look different enough to actually be cause by the change in timings?
incorrect cas on S3 resume is a bug somewhere IIRC.
---
Yes, I can confirm that this does work as all I've done between the above links and this one is updated to bios 0601 and changed B2B from disabled (auto in bios) to 10.
http://i188.photobucket.com/albums/z...utweakers3.jpg
There is no allotted time you have to let it sleep. Just put it in S3 sleep and then wake it up and you can change the timings.
Hey guys, a couple of questions.
"This should fix turbo throttling issues" - what is the turbo throttling issue?
Also, I noticed that many of you have extremely low voltages on your CPU, it seems my C0 920 won't run stable at 200x19 unless I give it 1.381 volts. Everything else is pretty much auto. Is there anything that I need to do in order to reduce my core voltage to run at this speed or is this simply based on the chip that I have?
When you are using the 21X multiplier on a Core i7 920, you are automatically using the turbo mode of the processor.
On the Asus P6T based boards, when you start pushing them too hard, at full load, the multiplier will start to rapidly cycle back and forth from 21X to 20X. This is the Intel design spec but very few motherboards have bothered to follow the Intel guidelines. Some motherboards give you an option to disable this safety feature in the bios but the Asus P6T boards don't have this option which can be a problem for Xtreme enthusiasts.
If you are overclocking your board hard enough and like to see your multiplier locked on 21.0X even when running Prime95 and LinX with the temps going over 80C then you might be a candidate for one of the special bios versions that Asus has been nice enough to release to keep the XS guys happy. It will eliminate turbo throttling at the risk of damaging your CPU or your board. Most guys on XS don't mind a little extra risk if it allows them to run some bigger numbers. That's why this is called XtremeSystems.
This was a bigger problem for the original C0 processors since they tended to need more core voltage to remain stable which would lead to turbo throttling sooner.
I am having some weird issues trying to OC with 12GB on this board and was wondering if anyone could give me some advice:
i7 920 D0 @4.22Ghz
6x2GB Corsair DDR3 1600 (CM3X2G1600C9)
BIOS 407
Current stable BIOS settings:
Multi: 21x
Base Clock: 200
RAM: DDR3 1203
Vcore: 1.35v
vQPI: 1.65v
VDimm: Auto
If I set my vDimm off auto, I only detect either 8GB or 10GB of the RAM at boot up depending on the other settings. If I set the RAM speed at anything above the lowest value, I can't get more than 8GB to detect on boot up. I've tried manual vDimm values from 1.5v all the way to 1.7v, but it still won't detect all of my RAM like the auto setting does. I've tried using both the XMP profile and straight manual overclocking, but it doesn't seem to affect the result.
Any ideas on what else I can try?
i7 920 @ 4.2Ghz
Asus P6T6 Revolution
3x GTX260
6x2GB Corsair DDR3-1600
G.Skill Falcon 128GB SSD
G.SKill Titan 128GB SSD
Segate 7200.11 1.5TB
Vista 64 Ultimate
Seeing less than installed RAM usually means the RAM is not getting enough voltage or the timings are too tight for the speed you're running at or a combination of both. You should set your VDIMM to 1.64V to start. You can use higher than 1.65 just up your VTT from default. There should be no more than 500mV differential between the two or you may have issues. A 1.3V VTT while still modest would support a 1.8VDIMM.
EDIT: Why is your QPI voltage so high?
You need to set vdimm manually to a value within 0.1v to 0.5v of QPI/DRAM. As to the value, through trial and error you'll find the optimal value for vdimm. If that doesn't work, then you need to start raising QPI/DRAM and VDIMM together. I say your QPI/DRAM voltage is too high as well, start from 1.35v QPI/DRAM and 1.66v VDIMM and work you way up. But test each mem module in each slot as well.
Possible reasons for detecting less memory than installed http://i4memory.com/f54/asus-x58-6-1...nstalled-19200
1. incorrect cpu vtt (uncore) voltage set - what Asus terms as QPI/DRAM voltage
2. incorrect QPI/DRAM to VDIMM relationship - every cpu/mem pairing will have an optimal voltage differential between QPI/DRAM to VDIMM voltage usually between 0.01 to 0.5v difference. Fall out of the optimal differential voltage for cpu/mem pairing will = less memory detected and/or lower memory bandwidth reported compared to when optimal voltage differential is in place.
Example, say you have 6x2GB @1333mhz working fine at 1.35v QPI/DRAM with 1.65v VDIMM. You try to overclock the memory to @1600mhz speeds, and you bump VDIMM from 1.65v to 1.75v and now only see 8-10GB detected. You've moved out of the optimal differential range for QPI/DRAM to VDIMM which @1333mhz was 1.65 - 1.35v = 0.3v. At 1.75v VDIMM you left QPI/DRAM at 1.35v so 0.4v differential voltage. Then you bump QPI/DRAM to 1.45v to keep 0.3v differential voltage and 12GB is detected again. Now this is just an example, it could be you need 1.85v VDIMM for your memory to be stable and if 0.3v is optimal differential voltage it means you need QPI/DRAM set at 1.55v
3. faulty memory modules
4. faulty memory dimm slots
5. sort of combination of 3+4 above where memory modules like particular memory dimm slots - could come down to PCB differences for each module along with voltages for QPI/DRAM and VDIMM.
Also don't count out needing to tweak, IOH, IOH/ICH PCI-E, ICH voltages as well as DRAM DATA/CTRL REF voltage tweaks for each memory Channel A/B/C
When testing full memory dimm slot populated configurations, it's also worth testing each memory module (label them with masking tape from #1 to #6) in each of the 6 dimm slots, jot down notes as to how each module performs in each of the 6 dimm slots, then pair them up with best matching of memory modules for dimm slots. Reason is triple channel kits while sold to you as matched, they are rarely really matched in terms of overclocking characteristics - you'll always have a module or two which either does better or worse that the average of the 3 modules. Now take that to 6 modules and you can see why sometimes it might not work that well.
HTH
George
Last edited by eva2000; 09-16-2009 at 06:13 PM.
---
I also get this problem when going for 4.4ghz on my board and 12gb of Dominator 1600mhz RAM, though i'm trying to overclock the RAM as well. It happens when set to the lower multi for the RAM. Tried VTT all way upto 1.45v but still would only detect 10GB RAM. Guess need to up the RAM volts as well to stick in the "zone"
Have found that quite often too much voltage is just as bad as not enough, need to hit that sweet spot for the overclock to work.
Also strangely when testing for 4.4ghz with 6gb of RAM it was failing (BSOD on windows load) even when adjusting the voltages upto 1.45v for both Vcore and VTT, tried IOH, PCI and CPU/IOH skews - made no difference. Put the other 6gb back in and it booted, with 12gb RAM and ran superpi at 1.375v for both vcore and vtt!
It certainly needs some tweaking for higher clocks, but for quick clocking to 4ghz its certainly very easy to do.
i5-2500k @ 4.6ghz | Z68X-UD3H-Gen3 | 8GB Ram | EVGA HD GTX570
I can't go over 205 BLCK without crash in Win. Any advice?
Anybody sells DOMINATOR GT, 6GB (3x2GB), 2000MHz, 7-8-7-20?
Please contact me if you want to sell them: I really need them.
Rambo87 sent me a good looking screen shot that I forwarded to Asus. The multiplier is rock solid at 21.000 on all threads with special bios 0001 even when pushed to 4.4 GHz.
Good work Rambo87
Depends on the error you get on the BSOD as to what you need to do, grab that and we should be able to help
After a lot of testing, I found that when going for 4.4ghz i was needing around 1.375vtt, Vdimm at 1.64v, if i nudged vtt to 1.38125v then it would start detecting 10gb RAM, i nudged Vdimm to 1.66v and back to gettin 12gb recognised. Very nearly got it stable at that but temps hitting 81-82c on hottest core.
Unclewebb - I'm also on 0001 bios so will retest with the turbo throttling tool and see if mine jumps around as well, as i'm close to the temps and volts you said it happens (80c and 1.4v)
i5-2500k @ 4.6ghz | Z68X-UD3H-Gen3 | 8GB Ram | EVGA HD GTX570
Any new bioses for now?
no new bioses yet newest is still 0001 8/14 dated one
clocks memory nicely this board
New personal best cpuz memory validation and first time breaking DDR3-2300Mhz for triple channel on X58 @DDR3-2305Mhz 9-10-9-24 1T at 1.78v. Previous best was @DDR3-2298Mhz 9-9-9-24 with 3GB Samsung HCF0 modules.
System:
- Intel Xeon W3540 3845B010
- EK Supreme LT 1366 / FESER 480 / MCP655
- Asus P6T6 WS Revolution
- Gskill Trident F3-16000CL9T-6GBTD
- 512MB Inno3D 8800GT OC Edition 190.62WHQL
- 750GB Samsung SATA
- Pioneer DVR-217 DVD-RW
- Corsair HX1000 PSU
- Dual Boot - WinXP Pro SP3 (nLite fully updated) with Vista Theme & Win7 Ultimate 64bit RTM
Max Validation @DDR3-2305Mhz 9-10-9-24 1T at 1.78v
QPI/DRAM = 1.65625v
VDIMM = 1.78v
Uncore frequency = 4608Mhz
Super Pi 2M @DDR3-2256Mhz 9-10-9-24 1T at 1.78v
Max Super Pi 32M @DDR3-2220Mhz 9-10-9-24 1T at 1.76v
---
nope i7 turbo reports calculated multiplier under load as being 21x which is correct.. = no throttling as i understand it
---
Where can we download Bios 0001 ?
BulldogPO: You need to upgrade to RealTemp 3.30.
http://www.fileden.com/files/2008/3/...alTempBeta.zip
RealTemp 3.00 used to reset / restart the system timers before they wrapped around which in hindsight, wasn't a great idea. Since then I've learned to just let the timers run and when they wrap around, deal with it but don't ever try and stop them or predict when they are going to wrap around. When testing with RealTemp 3.30 and i7 Turbo 6.94, your multiplier should be very steady at 21.000 and it looks like it is. If you were getting a single glitch about once every 5 minutes, it was likely being caused by RealTemp 3.00. I'll take half the blame but Intel gets the other half for using 40 bit timers that wrap around way too frequently at full load. Most internal timers are 64 bits wide that take years to wrap around but not these ones.
Edit: Thanks eva2000 for posting the link, you beat me to it.
Last edited by unclewebb; 10-09-2009 at 08:19 AM.
Bottom of 1st post at http://i4memory.com/showthread.php?t=18752 or first post of this thread
sorry if this has been asked before but what settings should I adjust to fix the cpu over volting problem? I cannot set below 1.25v for the cpu in the bios and when i set that, the actual voltage is 1.4...
Thanks!
Bookmarks