Mine still drop in both LinX and Prime95 at 4.40 with HT enable when load temperature approaching 80'C, if I can keep it under 75'C then it remain at 21x.Quote:
Originally Posted by 003
Printable View
Mine still drop in both LinX and Prime95 at 4.40 with HT enable when load temperature approaching 80'C, if I can keep it under 75'C then it remain at 21x.Quote:
Originally Posted by 003
Mind to share your settings?Quote:
Originally Posted by Zucker2k
I am still "learning" this board (P6T) and so far, cannot get it stable with BCLK above 180.
Temp. is rediculous..... Currently, have my i7 920 running at 3759MHz with max loaded temp on 1 core is 88*C (other 3 cores' are around 83*C-84*C) - with watercooling
I think CPU TM Function works and throttle the cpu down when cpu reaches some limit temperature. ( but i thought we can enable o desable it)Quote:
Originally Posted by chuckbam
I´ve never had my cpu above 77ºC.
The only time i saw my cpu below its x20 multiplier was at default bios setting, when idle multy goes down to x12.
But since i disabled CPU TM Function and C1E the cpu is always at x21, may be ´cos never gets more than 77ºC.
I thought Disabling : SpeedStep, C1E and TM Function will never throttle down.
Hmm mine will throttle down at a way lower temperature. What voltage are you using?Quote:
Originally Posted by ben805
I think you are right. With the CPU TM Function disabled you should not get a throttle down. My self, I would keep TM Function enabled. And if you do get a throttle down, the fix would be at your cooling. I think it is like your dummy oil light on your car. It is not to limit your OC, but is just a safety switch.Quote:
Originally Posted by martinakerman
I maxed out everything today with LinX + Prime95 + two loops of 3DMark06 (simultaneously) and had no throttle down at a MAX CPU temp of 76C. One day I am going to see if I can get it to do it. But, maybe it would just blue screen.
I fixed a computer with a Core2 that was throttling down. Big wad of cat hair in the stock cooler.
I guess if you fire that system up an turn it off when done you have your hands on the controls. But, many people keep their computer up for days.
Mine still throttles, even with TM Function disabled. What needs to be disabled is the OCP limit on Core i7 which the P6T Deluxe does not allow you to do. The Rampage II boards do allow it, as do all DFI, gigabyte and evga boards afaik. Major fail on Asus' part.
How hot is your CPU getting? Maybe you need a cooler change? What is your Vcore?Quote:
Originally Posted by 003
I ran this all Simultaneously and no crash or throttle down. I maxed out my 12 GBs of RAM Man. I only have Air Cooling so I take it EZ.
http://chuckbam.com/i7_P6T-D/TT051409.PNG
Its turboboost throttling not what you may be thinking. It's really not going to be an issue for probably 99% of the rigs being used since you have to be pushing decent clocks, voltage and temps under extreme load for it to happen.Quote:
Originally Posted by chuckbam
Ah, if you are at a lower voltage and speed due to air cooling I don't think you will notice the problem. It is related to power usage, not heat (although heat can be a factor when the power is too high).Quote:
Originally Posted by chuckbam
For example if I am at 200x21 with 1.49vcore, I don't see temps over say 80C, but if I run LinX the multiplier instantly drops to 20x.
http://www.fileden.com/files/2008/3/3/1794507/Turbo.zip
I wrote this little program to try and keep a closer eye on the multi. At full load when Turbo mode first starts to cycle on and off, some tools might not accurately report your multiplier. The i7 Turbo tool uses high performance timers within Intel Core i7 CPUs to accurately calculate the multiplier at full load.
http://img268.imageshack.us/img268/7120/multi.png
I was not 100% sure how that worked. And really still not. The Specs say 130W and now they came out with a thermal for the 920 of 67.9C. I used to the thermal days.Quote:
Originally Posted by 003
Personally, I really don't like the (Turbo Mode) 21x setting. I like it more straight forward. Can you use the 20x and a little greater BCLK? Or, does that impact your memory?
OK, this is a Rampage II Extreme thread. But, they are not running a Vcore as high as you are at like OC. You may want to look at this thread. Same CPU, and I think Power Phase.
http://www.abxzone.com/forums/f41/ra...-114705-5.html
link does not seem to wrk??? h--p://www.abxzone.com/forums/f41/rampage-ii-extreme-114705-6.html
New bios for Asus P6T Deluxe V2 - P6T-ASUS-DELUXE-V2-0502.zip
Personlly I would perfer the 20x multiplier, but, it is way more unstable than 21x. That is why I am pissed off about this. And 21x would be every bit as stable as 20x if asus would enable the option to disable the TDC limits on the P6T Deluxe. Again, all the Rampage boards can do it, as well as all gigabyte, dfi and evga boards. Asus is the only one that restricts this option to their high end expensive boards. It's really starting to piss me off.
uncle:
That is a very good program and should demonstrate how unstable 21x really is on the p6t deluxe because of asus's limitations.
Real neat program. What does the rest do? Nothing at OpenLibSys.org. Does it just reset the counter not the hardware? Also, it seems to me that if you have turbo disabled in the bios the check box does not work.Quote:
Originally Posted by unclewebb
http://chuckbam.com/Post2/tCapture.PNG
are the ocz blade 2000 cas9 working on this motherboard? I am going to buy those rams......
working ...I ever try :):)Quote:
Originally Posted by okorop
I'm working on eliminating EIST kicking in at a certain voltage/temp. In my experience, voltage seems to be the culprit. At 1.344v and 4510mhz, I get EIST kick-in at around 71c. But with 1.336v and the same 4510mhz, I get no EIST kick-in even at 74c.
For those trying to boot at 21x220 on the vanilla, try these:
pcie freq: 104-108
qpi/dram: 1.4 - 1.425v
ioh: 1.28 - 1.30 (very important in my case)
cpu amplitude is at 700 for all my stability tests.
A little stablity run at 4.4Ghz, I managed to drop volts a little bit.... so far so good.
http://www.hostthenpost.org/uploads/...18cc47bc61.jpg
YEAH!!!!!!!! TURBO BABY !!!:shocked:
http://farm3.static.flickr.com/2261/...92f4d4a9_o.jpg
This is what i got after 1 hour of playing Call of Duty 5, i left this I7 Turbo program to see if my cpu throtthles down....:eek:
The Turbo tool is designed to keep an eye on the multiplier and is very accurate and very stable when all 4 cores are fully loaded. If you are running LinX or Prime then a 920 with Turbo enabled should stay locked at 21.0000 like it shows in Zucker2K's screen shot.
At idle, the Intel algorithm that I'm using has a few issues and can result in some random values. They don't occur very often but they do happen so I might try to filter those out. Anything beyond 21.0 with 4 cores and Turbo enabled on a 920 is obviously a mistake and would be easy to filter. It might be an idea to log all data to a file so you can go back and have a look or take the data into Excel for easier analysis. A dropout or spike for 1 second is normal but a constant drop below 21.0000 at full load is a sign of the Turbo cycling on and off.
The value in the MSR box is the multiplier that other programs like Core Temp read and display for the multiplier. It may or may not warn you when Turbo cycling first starts to happen.
The Reset button only resets the minimum and maximum values.
The Turbo Mode Disable feature can only be used if you previously enabled Turbo in the bios. It has no effect if you have already disabled Turbo in the bios.
Here is the Intel documentation that this tool is based on.
http://download.intel.com/design/pro...ots/320354.pdf
The value displayed in the Multiplier box is the average multiplier for the previous 1 second interval based on Intel's algorithm.
I got "boot block in file is not valid!" errorQuote:
Originally Posted by angelreaper
1.40v @4.4Ghz with HT enable, or 1.46v @4.5Ghz with HT enable. Throttle kicked in when loading in both Prime and LinX, this doesn't seems to isolated to only LinX. I ran a series of cpu benchmarks from Everest at 4.51Ghz with HT enable and notice it score worst than 4.40Ghz, ran the monitor software and bench again, this really suck as the multiplier get throttle back down to 20 x215, my 4.51Ghz overclock practically became 4.30Ghz under heavy load...which is not good!! :mad:Quote:
Originally Posted by 003
Worked fine on my V2 save odd bootup issues until the bios was cleared and manually re-set. I also found that the G15 keyboard not being detected can be solved by setting the USB 2.0 speed to "Full" rather than "HiSpeed". Unfortunately, that lowers the USB speed from 480 to 12 Mbps......Quote:
Originally Posted by angelreaper
I dont think OCP is the a factor when my multi drops. Also, i think this program might be too sensitive. Take a look at my max multi.
I have disabled EIST and C-states in BIOS, but the tool shows its enabled. But thats because the 21x multi uses Intel Turbo wich again uses EIST??
Unable to upload pictures.... bah.
I will edit my post with picture as soon as I install FlashFXP and find my logon info.
Meanwhile:
4009Mhz - 191x21
1.296v - CM Hyper212
Max temp: 84/82/76/77
i7Turbo:
MSR 21.000
Min: 17.986
Max: 27.934
It definitely should not drop below 20 if you have it set to 21, if it is, that is either a reporting error or you have something else going on, possibly EIST, C1E, C-State or CPU TM Function. unclewebb already said there were a few issues with i7 turbo such as the high multiplier reporting.
EIST and all the other powersaving features is disabled in BIOS.Quote:
Originally Posted by 003
When disabled in BIOS, is EIST suposed to show as enabled or disabled in the i7 Turbo app?
Im still running prime, and since I last posted the i7 Turbo new reports 15.9x min multi.
I have been staring at the the program, and when the multi changes its only for half a second before its back at 21x.
EDIT: According to the i7 Turbo app my multi changes every single second when idle. Min 20.8 and max 21.2. I dont get it.
EDIT2: I just read unclewebbs post, so never mind :)
Is there any other mobo that doesn't affect by this multiplier throttling crap? may be time to ditch this thing... :DQuote:
Originally Posted by 003
I hear you.Quote:
Originally Posted by ben805
DFI, Gigabyte, EVGA does not suffer from this, AFAIK.
Ya, the Asus Rampage boards (more expensive), and all the other boards from DFI, Gigabyte, evga (I think) etc.. not sure on MSI.Quote:
Originally Posted by ben805
Ya sorry, i meant to say Micron instead of Marvell.Quote:
Originally Posted by highoctane
And if anyone is interested, Asus has listed Windows 7 drivers for the P6T.
Well I have the V1 board and I was using it last Friday with no problems. And tell the gardner decided to cut my FIOS cable while he was trimming the trees and now I can not get my ethernet to work again. I tried everything, I think so far. I think it must of fried my 2 ports or something. I am using a second board to go online. Well now I have to setup a RMA through ASUS hope it goes well. Atleast I can throw in my new Bloodrage board for the time being. :(
Anyone has tried BIOS 0502 on P6T Deluxe V2? I had upgraded mine and unable to change mine DRAM Frequency setting except [Auto].
Yes:Quote:
Originally Posted by Niuminoid
I will try tonight 0502;)Quote:
Originally Posted by Truckchase!
Quote:
Originally Posted by angelreaper
Thanks for the feedback angelreaper :up:
Hope to hear from you very soon.
I'll most likely get the EVGA as replacement, not sure if I should get the A1 with 2yr warranty or spend more to get the classfied yet....is Asus going to fix the thermal throttling ever? before I get a replacement board I did a little comparison of my D0 in power consumption and here's what I found:Quote:
Originally Posted by 003
2.66Ghz @1.15v: 130w
4.00Ghz @1.22v: 219w
4.20Ghz @1.30v: 262w (underload throttling back to around 20.75x 200; 4.15Ghz)
4.30Ghz @1.35v: 289w (underload throttling back to around 20.5x 205; 4.2Ghz)
4.40Ghz @1.40v: 318w (underload throttling back to 20x 210; 4.2Ghz)
4.51Ghz @1.45v: 349w (underload throttling back to 20x 215; 4.3Ghz)
note:CPU wattage calculation from http://extreme.outervision.com/PSUEngine
For 24/7 I guess the sweet spot of my D0 is somewhere around 4.2~4.3Ghz with the least amount of heat load.
Found no difference with the old bios:shrug:Quote:
Originally Posted by Niuminoid
To me with xeon is still bad:
In the bios missing:
DRAM Multipliers other than 1:3 and 1:4
QPI Transfer Rate
Uncore Frequency
Ocz blade 3x2 gb 2000mhz cas9 are working really well on this board. I have just tried for 20 minutes and here the screens:
1800mhz cas 7 stable with 1.66 vcore
http://www.pctunerup.com/up/results/...35_Capture.jpg
And then 2000 mhz cas9 cas8 1.65volts
http://forum.hwproject.net/imagehost...1513203c5f.png
I tried also 2000 cas8, I will post the screen later :)
Hmm...What about the DRAM i did mention earlier?Quote:
Originally Posted by angelreaper
hows yours? any prob with that? can u set the frequency?
Yes I can set the frequency.Quote:
Originally Posted by Niuminoid
Hmm..Strange..mine wont post after i've changed the frequency to 1333MHz.Quote:
Originally Posted by angelreaper
May i know what is your DRAM frequency? Are you using KVR?
1531. I can set also 1149 and AutoQuote:
Originally Posted by Niuminoid
http://img514.imageshack.us/img514/5231/ram.jpg
Just want to ask a quick question.
It there thermal protection on the i7 or p6t deluxe v2 ?
I remember my old P6b deluxe has thermal protection (i.e, when the cpu reach a certain temp., it will buzz an alarm or shutdown the system. I dont remember which.)
Did you clear your bios settings and re-set everything?Quote:
Originally Posted by Niuminoid
(setting from one of the load slots won't work)
I have same cpu, mobo and ram than yours. my D0 is at 1.35625v @ 205 x 21 = 4.3Ghz but i have never seen it throttling back.Quote:
Originally Posted by ben805
I tried I7 Turbo program to check the throttling but never any issue.
My cpu is able to reach 80ºC on Air but it doesn´t throttle back, how do you check the throttling..?
I think it occurs when voltage is set at 1.4v or so. And make sure HT is enabled.Quote:
Originally Posted by martinakerman
Last night I did more testing, had the A/C cranking in a small room and brought down the ambient temperature to 16C, at 4.40Ghz with 1.40vcore it idle only around 32~33'C, underload in LinX were 66~67C on all 4 cores and the throttling nonsense still kicked in and multiplier dropped down to 20x. So this crap is definitely not temperature dependent. Maybe like 003 suspected....it was affected by the vcore in a certain range...
HT is enable.Quote:
Originally Posted by 003
Try with less V core, with this settings (that you already know) i got no throttle.Quote:
Originally Posted by ben805
Ai Overclock Tuner: Manual
CPU Ratio: x21
Intel Speed Step: DISABLE
BCLK Freq: 200
PCIE Feq:101
DRAM Freq: 1603MHZ
UCLK Freq: AUTO
QPI Link Data Rate: AUTO
DRAM Timing Control: 8-7-7-24
CPU Voltage: 1.35
CPU PLL: AUTO
QPI/DRAM Voltage: 1.35
IOH Voltage: 1.22
DRAM Voltage: 1.66
Load-Line Calibration Enable
CPU Spread Spectrum: Disable
PCIE Spread Spectrum: Disable
C1E Support: Disable
CPU TM Function: Disable
HyperThreating: Enable
(All the rest: Auto/Default)
martinakerman: Can you post a screen shot of the i7 Turbo tool and CPU-Z while running LinX? Your bios setting is 1.35625v but your actual core voltage is likely a little less than that.
Turbo throttling is mostly controlled by core voltage and then MHz. At full load, a 5% increase in core voltage should increase power consumption similar to what a 10% increase in frequency (MHz) would cause.
As the maximum TDP is reached, Turbo mode will start to cycle rapidly, hundreds of times a second, on and off. When you really start pushing things hard on the Asus P6T, Turbo mode will mostly remain off. Software that only looks at your multiplier once per second will initially miss what's going on inside your CPU when Turbo cycling first starts to happen.
I have a few ideas to make the i7 Turbo tool better but I'm still waiting for some more feedback. If enough people complain about this then maybe Asus will get off their butts and fix this issue and give users a way to turn off TDP based Turbo throttling like most other motherboards allow you to do.
Post some screen shots so I can have a look and send a link to Asus to get them to wake up and admit that there's a problem here.
ben805: A $15 Kill-a-Watt meter is a great thing to have in the tool box for tracking down issues like this. It should be very clear when power consumption at the wall reaches a certain level that Turbo mode starts to cycle.
Edit: Here's an example of Prime Large FFTs at 4200MHz and 1.392 volts being enough of a load to start Turbo cycling. CPU-Z is still choosing to ignore that something is going on. The core temp is only 63C so that's not the problem.
http://img40.imageshack.us/img40/3954/throttle.jpg
We know this is related to turbo and the TDC/TDP limit on the non-EE Core i7 CPUs. Here is an article detailing it:
http://www.tomshardware.com/forum/25...ore-overclocks
But, the thing is, the motherboard manufactures can have the TDC/TDP limit disabled via the BIOS. DFI, Gigabyte and eVGA include the option on all X58 boards. It is only asus that limits this function to their "high end" Rampage series X58 boards. :mad::mad:
This has absolutely nothing to do with the power capacity from the wall, which is fine by the way when I checked it last time.Quote:
Originally Posted by unclewebb
I realize that's not your problem. I suggested the Kill-a-Watt just as a different way to monitor throttling.Quote:
Originally Posted by ben805
As you load your CPU and increase the core voltage and the MHz, at full load, a Kill-a-Watt meter will accurately show you how much power is being consumed from the wall. When running LinX, etc. at the desktop in 2D, it will indirectly give you a good idea of how many watts your CPU is consuming.
In theory, the throttling point should correspond very closely with how many watts your CPU is consuming. No matter what software, MHz or core voltage you're using; you will likely find that the throttling point is going to occur at the same power draw from the wall. This is just a different way to prove what you already know in case Asus is still in denial.
ahh I see, where can I get such device from? any specific brand or model? and indeed it would be very nice to have one, in the past I use my APC UPS to monitor the total power consumption under load and idle but it died few months ago, haven't got a chance to fix it yet.Quote:
Originally Posted by unclewebb
There's always lots of them available on EBay. Just search for Kill-a-Watt. Many hardware stores have them or something similar.
I used it when comparing core temperatures vs power consumption. It shows Prime95 Small FFTs having a very steady load which should result in consistent temperatures from core to core. LinX has a higher peak load which causes higher peak temps but with LinX, the load and power consumption drops significantly between each round. That's why I use Small FFTs when comparing temperatures from core to core and use LinX for creating maximum heat. I also used it to monitor and adjust my Sapphire 1950XT. Its bios was giving it more core voltage than it needed which was easy to see by watching the increased power consumption in 3D and explained the excessive heat. I have a hunch that the Kill-a-Watt would clearly show Turbo throttling at the same time that the i7 Turbo tool does.
Just another way to monitor what's going on with this board at full load. I think Turbo throttling is an issue on a few of the less popular boards from smaller manufacturers too but with CPU-Z sometimes covering it up, I'm sure many users don't realize what's going on.
Hmmm..you mean clear CMOS? Haven try it yet. Only tried load default settings and it doesnt work :(Quote:
Originally Posted by Truckchase!
For laughs I decided to do what you did initially and was able to replicate your issue. I followed the instructions in the manual and I unplugged the cord to the PSU, moved the CMOS Clear jumper over for 20 seconds, then moved it back, booted up and the issue is gone. Have you tried the setting in the BIOS that says "Load Setup Defaults"?Quote:
Originally Posted by Ozan Baseski
Also, try using a different jumper. Maybe the contacts aren't being made within the jumper. To be sure you put the jumper back correctly--you have the jumper over the two pins that are right next to each other--correct?
Here it isQuote:
Originally Posted by unclewebb
http://farm4.static.flickr.com/3028/...1d404fa6_b.jpg
This screenshot was taken with this bios setup:
Ai Overclock Tuner: Manual
CPU Ratio: x21
Intel Speed Step: DISABLE
BCLK Freq: 200
PCIE Feq:101
DRAM Freq: 1603MHZ
UCLK Freq: AUTO
QPI Link Data Rate: AUTO
DRAM Timing Control: 8-7-7-24
CPU Voltage: 1.35
CPU PLL: AUTO
QPI/DRAM Voltage: 1.35
IOH Voltage: 1.22
DRAM Voltage: 1.66
Load-Line Calibration Enable
CPU Spread Spectrum: Disable
PCIE Spread Spectrum: Disable
C1E Support: Disable
CPU TM Function: Disable
HyperThreating: Enable
(All the rest: Auto/Default)
Yes I tried everything including loading setup defaults and flashing new bios. Anyway last week I sent the board for RMA. Today they contacted me and told me that it was fixed. They had replaced a transistor near the "Chassis Intrusion" jumper which has been damaged when I removed the wrong jumper. It's really interesting how a board can stop working just by removing and putting back a jumper. Anyway I'll get my board tomorrow and get you guys informed about the situation.Quote:
Originally Posted by Burner27
Btw I am also thinking about replacing this board with a new one. Which brand would you suggest? Among my choices are new Asus P6T Dlx V2 (mine was V1), Rampage II Extreme, Giga UD5-UD4P, MSI Eclipse, DFI X58 T3eH6.
P6T DELUXE-1604 bios has been removed from the ASUS websites.
You're going to replace a perfectly working returned board because of the simple jumper issue? Well, it's your choice. Frankly, I think jumpers should not be anywhere on a motherboards nowadays. If you're really intent on switching to another brand, I would look at the eVGA X58 motherboard, or the Gigabyte UD5 x58 motherboard as potential replacements. The P6T V2 is different from the V1 in that is lacks the SAS ports and has jumpers; the RIIE has jumpers, the Gigabyte UD5 may be a worthy choice; the MSI got good reviews from MaximumPC Mag; don't know anything about the DFI. What makes you think if you switch to another board you will have any less issues? Why are you condemning a perfectly working board (when you get it back)?Quote:
Originally Posted by Ozan Baseski
Hmm why?? I've been running it and it is working great for me..Quote:
Originally Posted by chuckbam
I have been running too. Maybe a new one will be out soon.Quote:
Originally Posted by 003
P6T-DELUXE-V2-0502 bios has been removed also from the ASUS websites and ftp site:shrug:
Wow never knew 0502 was out to begin with...what changes was there fromQuote:
Originally Posted by angelreaper
0406 ,cause I'm still running 0406..
[ 0502 ]Quote:
Originally Posted by nge769
----------------------
Update Express Gate
----------------------
[ 1504 ] V1
----------------------
1. Fixed the key may lag in BIOS setup when a Turbo Mode ratio was selected.
2. Supported EIST function when CPU ratio is in manual mode.
3. Adjusted max value of CPU ratio setting according to the number of cores active in BIOS setup.
4. Fixed setup item "CPU Ratio Setting" may disappear if down CPU ratio.
5. Fixed Vista may loading slowly when populated memory size is larger than 3GB and with high-end VGAs
502 kinda sucks, I found it not as stable as the 406 and I lost both OC Profiles that i had saved, after had it for a day i immediately roll back to 406.
Ya the 602 bios for P6T standard has been removed too. I guess they found some issues
I tried to talk with Asus customer service about the throttling issue and got a very official message from them regarding how they don't support overclocking and how they can't help me. How I wish evga board didn't have the NB fan on board and didn't suffer from heat issues!!...:(
Did you get the evga board? How is the NB fan and heat on NB? I have heard its bad from many folks, but some folks say its alright. What was the case for you?Quote:
Originally Posted by ben805
Well yesterday I got my board from RMA.But the "Chassis Intrusion" problem is still there. When I look closer to the chassis intrusion jumper I realized that a transistor just below this jumper have been removed and its place is empty. Perhaps they couldn't solder it perfectly when they fixed it and now the transistor's place is empty. I think that's why the board is still giving the same error. Anyway I am really fed up with this board and I'll send it for RMA second time. I am considering of getting a Asus Rampage II Extreme or Gigabyte X58 Extreme. Evga X58 can't be found in my country. I can get it from abroad but then its warranty will not be valid here. My brand choices are Asus, Gigabyte, DFI or MSI.
New bios for Asus P6T Deluxe V2 - 0504
ftp://ftp.asus.com.tw/pub/asus/mb/LG...XE-V2-0504.zip
[ 0504 ]
----------------------
1. Support new Express Gate version
2. Fix system may hang after resume from S3
I hope for good support of Xeon:)
Please some one with V2 and Xeon to share how is the situation.
Thanks
Hey reaper, isn't that a V2 and a Xeon in your sig? :slap:
Yes but I`m far far away from home and V2 and Xeon:)Quote:
Originally Posted by r1rhyder
For me, nothing has changed. MC still locked for CPU-Tweaker too. Same vcore and QPI/Dram. Don't expect any improvements.
Bios 0603 for P6T standard
ftp://ftp.asus.com.tw/pub/ASUS/mb/LG...-ASUS-0603.zip
With Хeon you try?Quote:
Originally Posted by r1rhyder
martinakerman,
Why are u using CAS 8-7-7 and not 7-7-7? Your sticks are CAS 7 rated, aren't?
Again no full support for Xeon in bios 0504:shakes: It`s a shame:down:
I think the multi-dropping is triggered when the tcase (shown on Asus boards as CPUTIN under hardware monitor) exceeds 67.9. I watched realtemp and hardware monitor for quite sometime to conclude that. I think CPU is throttling because it thinks its in a hostile environment.
So, no matter what your volts are or what your ambient is, a consistent tcase greater than equal to 68C will drop the multi. Anybody can confirm that? Or not. How do we know sensors are working consistently across motherboards?
Asus needs to disable th damn throttling....:( Its so annoying to test something and use something else. I test 4.1 in Linx when I think it is 4.3Ghz....:( And it will randomly crash on me someday.
Asus bios for now is a really mess:down:
We urgently need a beta bios as Rampage II Extreme:
http://www.xtremesystems.org/forums/...60&postcount=4
I just finished a new version of my tool to check for Turbo throttling and started a new thread here:
http://www.xtremesystems.org/forums/...66#post3808966
Hopefully some P6T users can head there and post some results of their testing. I plan to start a list there of motherboards that have the Turbo throttling issue. If there is enough negative press then maybe Asus will fix this issue.
Probably not but at least a list might help keep users from buying a board that has a serious issue that the manufacturer is not willing to fix.
I could really use some of your results ben805 to show the world what this problem is all about.
There is a new bios 0504 in the ftp server for the v2 board.
And it doesn't have the fix for S3 resume (like it claims) and it doesn't have the fix for multi dropping from 21x to 20x. Such a shame!Quote:
Originally Posted by cheungtsw
Quote:
Originally Posted by unclewebb
Sorry for the late reply, have been busy doing a few photo gigs but whenever I have a chance I'll revisit this problem again soon. As long as Asus ignore this multiplier dropping issue I can not recommend this system board to anyone except to those who own a weak chip that are not Prime/LinX stable above 4.2Ghz. :stick: My finding so far is this damn board is handicapping my D0 chip to 4.2~4.3Ghz (@4.4~4.5 OC).
I still can't decide to get the EVGA vanilla or the Classified. Yea from what i heard the vanilla version has loud NB fan and they tend to run much hotter than the Asus which I have never seen rise above 45C under heavy load.Quote:
Originally Posted by devsk
I was in the same predicament , between the classified and the RIIE and i went with the RIIE and a extra gtx295 to boot ..cause I'm used to the bios more and the north bridge on the classified runs to hot for my liking. i should receive it tomorrow.;)Quote:
Originally Posted by ben805
Can you find out if there is a linkage between the tcase and multi dropping? tcase is reported by hardware monitor (winbond chip reports it as cputin). I think its not the TDP which decides the throttle but the tcase (which has a relation to TDP but is determined by the ability of the cooler to get rid of the heat from CPU's vicinity quickly), which when passes tcasemax, cpu throttles.Quote:
Originally Posted by ben805
vanilla and classified are $100 apart....:D Its not a choice for me. Although EVGA support may be good, who wants to takes the hassle of returning a fried board. And with those temps running in 90C's, any serious use is going to decrease the life of the board. I am surprised that EVGA overlooked this. I have a feeling that evga boards are designed with WC in mind.Quote:
I still can't decide to get the EVGA vanilla or the Classified. Yea from what i heard the vanilla version has loud NB fan and they tend to run much hotter than the Asus which I have never seen rise above 45C under heavy load.
And S3 sleep doesn't work on vanilla evga or asus boards, so there is nothing to choose there. If Asus can fix this multi issue, I am happy with sticking with Asus because this board is solid otherwise.
Yes, they are.Quote:
Originally Posted by McBacker
But at 7-7-7 @ 1.66v it makes the system restart , i should increase a bit more the voltage, and i don´t want to do that.
Other thing is i am running 12Gb the all 6 slots copulated, and may be this has something to do with it.
Ok... got it... i also own a CAS 7 rated KIT (Ocz Platinum) and it only runs at CAS 7 if i put 1.7v on them... anything lower than this will fail on Linx in the first 10 seconds... i asked about this at OCZ Forum and they told me it's ok to use 1.7v.Quote:
Originally Posted by martinakerman
Regards
The OCZ Platinum XTC 7-7-7 1.65v ?Quote:
Originally Posted by McBacker
I have 2x3GB of XTC in my Dlx V2. Tho I havent tested LinX, I can run prime just fine @ 1600+ Mhz 7-7-7 with 1.6v
Running 01 and spi @ 1750 7-7-7 with 1.66v
I have not yet maxed them out since my "old" C0 tops out at 4570Mhz in 01 and spi.
Im on 0406.
Are OCZ selling sticks with poorly binned chips? If the sticks wont do stock speed on stock vdimm, I would RMA them.
Some one crossflash P6T Deluxe V1/V2 to RAMPAGE II EXTREME? I see no another way to good support for Xeon:shrug:
I'm pretty sure I read somewhere else that if you are running RAM sticks in all 6 slots, on some boards you have to raise from 1.65v to 1.70v due to drop in voltage when running all RAM slots.Quote:
Originally Posted by Formann
I really want to know this as well...Quote:
Originally Posted by angelreaper
Formann,
Yes. OCZ Platinum XTC 7-7-7 1.65v ( 3 x 2GB )
RMA would take a lot of time...
Can u try Linx and report back?
Also, can u post the BIOS settings u are using? IOH, ICH, VTT, QPI/DRAM, VCore...
Thanks.
LinX 0.5.9 failed after ~1min with anything less than 1.7vQuote:
Originally Posted by McBacker
I ran LinX @ 3400Mhz since my Hyper212 takes a beating in LinX.
1.20 vCore
1.31 QPI/DRAM
1.20 IHO
1.20 ICH
VTT on i7? Did I miss something?
Somehow I dont think OCZ are using LinX/Linpack when they are testing RAM.
Or... maybe the mobo undervolts? Asus Probe doesn´t report vDimm, so its hard to say.
Maybe we should check with a multimeter.
In light of this, maybe it was the RAM failing @ 219x21, and not the CPU. Sure would be nice with another 100Mhz for benching :D
Formann,
Thanks for reporting!
Good to know I'm not alone on this =]
Btw, VTT is CPU PLL on P6T.
Quote:
Originally Posted by unclewebb
When I was looking for vDimm reporting in Asus Probe I found something strange that might be a factor in the throttling issue.
Asus probe reports CPU temp to be 8c higher than core0 in RealTemp 3.0
RT: 72/73/67/66
Probe: 80c
It wouldn´t be too far fetched to think that the "throttling monitor" is using the same sensor reporting as Asus Prome.
If I get the time i´ll do some testing later on. Maybe some of the throttling we´re seeing is because of the mobo thinking temps are exceeding 90/95/100c ?
bump up your QPI to 1.35 and keep vdimm at 1.66 and if they still failed, test one stick at a time and see what happen. few weeks ago I was pulling my hair out trying to figure out why I coulnd't get a post with 12GB, only to find out my P6T v2 board was defective to begin with, you can see my post at OCZ forum here: http://www.ocztechnologyforum.com/fo...ad.php?t=55657Quote:
Originally Posted by Formann
I have tried all the way up to 1.4 vQPI.Quote:
Originally Posted by ben805
Testing the sticks one by one is too much hassle. I have no trouble other than the sticks needing 1.7v setting in BIOS, and thats only for 100% linx stability. Prime and all other programs are fine even if I run 1650-1700+Mhz 7-7-7.
Who knows, maybe 1.7vDimm in BIOS is undervolted to 1.6v
I assume Asus would make absolutely sure its not overvolting, so a little undervolting would not surprise me.
TurboV reports vDimm, Dram reference, QPI, CPU, CPU PLL, ICH PCIE, ICH, IOH etc etc. I dont use it for overclocking but for checking voltages its great.Quote:
Originally Posted by Formann
Got the P6T standard but its the same on deluxe.
How do you get it to display actual voltages?Quote:
Originally Posted by Buggalugs
My TurboV only show the settings.
hi all,
i'm a new in this field so forgive me my qustions.
With the default speed of the CPU i can put my ram to 2000Mhz.
but with this current speed the max of my ram speed is 1683.
the next speed is 2000 windows does not boot!
i don't know what i have to change to make it work
any help?
and also any advice to tweak these setting, i will be glade
i7 920 D0 Batch 3849A866
Asus P6T Deluxe V2 buis 406
RAM 6GB OCZ Blade DDR3 2000 cl9,
Code:CPU Ratio : 19
SpeedStep : Disable
BCLK : 210
PCIE : 101
DRAM: 1683
UCLK : Auto
QPI : Auto
DRAM Timing : Auto
CPU Volt : 1.23125
CPU PPL : 1.96
QPI/DRAM : 1.36875
IOH : 1.22
DRAM Bus : 1.66
CPU Differential Amp : 800mV
left is AUTO
C1E sUPPORT : eNABLED
HT : Enabled
C-STATE : Enabled
C Sate : Auto