I guess there is no ETA for that fix.. just expect it in the next bios update?Quote:
Originally Posted by chew*
Printable View
I guess there is no ETA for that fix.. just expect it in the next bios update?Quote:
Originally Posted by chew*
No ETA but they are aware.....
I have the 790x-ud4 and i have the same problem as the rest with the nb-multi :)
A quick question, is there a easy way to remove the nb heatsink? Ive only looked at it fast but is it glued or how is it attached to the mb?
edit: Nevermind my question, just saw the pushpins :)
But is it worth going the wc road for the nb? I have the block anyway so its basicly just sitting here waiting. Then i guess i remove the heatpipe and the heatsink and leave the heatsink over vrms. The vrms should be colder after or? Since they just handle their own heat and not their own + the nb.
Ohhh i can wait. this Mobo has lot of Potential.Quote:
Originally Posted by chew*
Remove the two plastic pins at each end and gently wiggle to remove the Sinks.Quote:
Originally Posted by Snowman89
NOTE:
If the Upper Heatpipecooler is attached to the NB(middle heatsink) you will have four pins total and need to remove the unit as one piece.
I usually change the paste on New Mobos.. This is the first one i have not . I want to test the 550BE once it gets here since the Mobo is in a case. Temps seem to be very resonable on my MOBO with Stock heatsinks(blue Shields included)seeing i have a lot of Xflow in my case(2x120's going accross Mobo:D).
But cant i remove the heatpipe first or how is it attached to the heatsinks?Quote:
Originally Posted by gOtVoltage
ud5p doesnt have this problem, does it? (making a new build for a friend ^^ ) Also with this board, does Gskill have have any issues, or on the XT version with ddr3?Quote:
Originally Posted by chew*
YOU need to cool the VRMs. If you water cool the NB Make shure you keep some type of heatsink ON the VRMs ,,Quote:
Originally Posted by Snowman89
Thats why there is a BiG A$sheatsink on them.
I unsolder the Heatpipe at the VRM end and keep the VRM heatsink on mobo. Then throw pipe and NB in the box incase of the inevitable when using W/B on NB and only cooling NB with water.
Some people unsolder the heatpipe instead of cutting so you can resolder it and NOT(void warranty) on RMA if need be.
Just Heat and you can slide back in,, if RMA knocks at the Front Door.
guys just set the voltage........Apparenlty it appears the CPU_NB vid is working.....however not detecting voltage correctly.....
So if its @ 1.2vid and we ADD .250vid this should = Close to 1.45vid?Quote:
Originally Posted by chew*
been leaving it stock @1.2vid because nothing shows what its really @ when you ADD to it.
Tomorrow ill check it out again once i installl the 550BE.
. it would be nice if showed the Volts next to it like Vcore and Vdimm does.
Yah that would be nice but, I pick my battles.....working board over software or voltage monitoring.
What software are you using to monitor the voltages?Quote:
Originally Posted by gOtVoltage
You can use (HWmonitor) and read TMPIN0 is chipset,, TMPIN1 for CPU,,TMPIN2 is PWM on this Mobo.Quote:
Originally Posted by Snowman89
http://www.cpuid.com/hwmonitor.php
You can also us (Everest) in the Overclock menu and select sensors.
It will Show the VIDs and other voltages. but wont read temps correct.
Something seems really fishy about that statement to me chew*.Quote:
Originally Posted by chew*
I'm not basing this on what voltage monitoring software says, I'm basing it on the way the system reacts. And it acts like the IMCv's are too low, plus the fact that when you change it in Windows all the monitoring software shows the changes so I don't think it's a detection problem.
I don't have time right now, but when I get home from work I'll try to run some tests that demonstrate what I'm talking about.
Looks like GBT may be up to they're old tricks of just denying a problem.
The core temps have been ridiculously low since the day I built it.Quote:
Originally Posted by Daveburt714
I don't currently have the 4th core unlocked. I tried it with BIOS F2B and unlocked it successfully but then the audio was all messed up. I needed a working PC and didn't have time to figure out the audio thing so I flashed back to F1 and then to F3 (what I'm on now) later on. Haven't tried enabling the 4th core again and my core temps are still reporting too low.
Ahh, I thought they went TMPIN1, TMPIN2, TMPIN3 but on second look, you're correct... wasn't using my Phenom system when I posted.Quote:
Originally Posted by gOtVoltage
So the top one in the list, TMPIN0, is chipset? That's the one that was getting very hot before I removed the blue covers and put a 120mm fan in the side panel. The last one, TMPIN2 (you're saying it's PWM/VRM) barely fluctuates, so that makes me wonder if the heatpipe going up from the NB heatsink is defective or the heatsink isn't making good contact with the chip.
I wish Gigabyte would use nylon screws or the wire clips for mounting their heatsinks instead of those disposable white plastic clips. :rolleyes:
Well dave they told me same thing I told you...... Probe to verify......according to them voltage is correct.Quote:
Originally Posted by Daveburt714
I don't have a board so I can't do it.
I dont have a volt monitor, that i could use, but where on the mobo would you take the reading from to confirm?Quote:
Originally Posted by chew*
Also if the nb volts are working... I would also tend to agree with dave burt as to what i ve seen, i ve been rock solid up to 3.5ghz, 3 core/4cores doesnt matter, but anything above, i start to get page fault in non paged area bsods.. i have my nb at 2200mhz, I do have 8gb of valueish mushkin ram, but timings are loose, and not oc'd, If it were other bsod's it would make sense , my stability doesnt seem to fluctuate on nb vid.. with I have in bios at +.200v so an effective 1.4 amds safe max.. I ll play again systematically this weekend maybe and see what i can come up with, could be my nb is taxed at 2200?
Its not maxed,, I have Done 2400 and My NB does 2800+ @ 1.45vid on two other Mobos.Quote:
Originally Posted by villa1n
On THIS MOBO even with +.200,,,250,,.300 = 1.5vid it wont do crap on the NB .. i dont know hwer to look but there is something going on . It may be volts or even Ram related ..
Im still playing around with the Nbvid to see if i can see Temp fluctuations when i raise and lower it. So far i it dosnt matter what i do it seems to stay the Same.
This Same 720 can do 3700cpu/NB2800+ @ 1.55vid with no erors/bsod's on the M3A79_T
I know the M3A VId works 100%. You can see it in the Nb Speed at any speed.
Somethings happening to keep the NB @ 2400 and lower.
Ok guys, as promised I did some tests tonight and here are the results.
Conditions: I uninstalled PhenomMSR Tweaker before the test to avoid any potential interference. I was going to continue the tests with just lowered CPU and Memory settings (to take everything else out of the equation). Unfortunately, with my NB Multi @ 12x (setting used with PMSRT) I got to the Win7 password screen, entered password and the machine failed to load windows... So I used the HTmulti to tweak as close as I could and still be bootable.
Heres the bios settings:
As you can see IMCv's are set to 1.30v
CLICKABLE:
http://i30.photobucket.com/albums/c3...v_BiosShot.jpg
After boot, all I did was open Everest, AMD Power Monitor and 2 CPU-z Screens. Both Everest and AMDPM show NBvid @ 1.20v...
Then I ran a 10 pass IBT run. The result was an "Unstable system warning", IBT failed!
CLICKABLE:
http://i30.photobucket.com/albums/c3...t_BiosOnly.png
Without making any other changes, I opened AOD and added .05v to the IMC making sure that all other settings and voltages remained the same, then closed AOD. All this was done without any other changes being made and without a reboot. I then ran another 10 pass IBT run.
CLICKABLE:
http://i30.photobucket.com/albums/c3..._AODplus05.png
Now, both Everest and AMDPM show the IMC at 1.25v and the IBT run passes succesfully. Maybe I'm crazy, but I don't know how else to explain this other than the fact that the monitoring software is working correctly and the bios is NOT setting the CPU-NB vid... :(
chew*: I may be able to borrow a DMM from a freind and it seems like the UD5P and UD4P are at least physically similar, could you shoot me a pic of the NBvid measurment points on the UD5P?
At least it would give me a starting point of where to look.... :yepp:
Sure after my PSU' s arrive tomorrow and I hook everything up to boot it.
Thanks Bro... :up:Quote:
Originally Posted by chew*
I finally got a reply from their online tech support. I don't know if its a language barrier or what, but it's as bad as I remembered with the DS5 and S2H....
Asking me the same questions I had explained in the previous message...:shakes:
This really is a MUCH better board than either of those, and I can work around the problem to some extent by using Phenom MSR Tweaker but this is a pretty serious problem.
I hope they don't just give up on it...
Hopefully the "powers that be" are watching this thread and can explain the results I got in the previous message... ;)
I tried MSR Tweaker too.. Now I just Installed AOD 3.2.. I hate AOD..for me to install and use it means theres something wrong lol.Quote:
Originally Posted by Daveburt714
the NBvid is controlable in AOD ONLY ,, I dont use AOD for overclocking so this sucks... After setting AOD and putting some stress on system i can see the higher temp loads from the ADDED NBvidvolts on the NB.
It seems the NBvid is NOT working in Bios. It can only be adjusted in AOD or Other software. Ive tried everything , Nada, even if i set it to the MAX Nada. CPUvcore is the only CPU volt setting we Have that worksand i have even set that to max. Yup it does Max on CPUvcore 1.81v:D
ONLY (Vcore is adjustable). I think teh GIGAguys think we may have been talking about the auto CPUvid or something instead of NBvid.This is so easy to see happen. NBvid does Not work period when trying to change in the BIOS.
O when i Raise the NBvid (USING AOD) , since the Bios does keep it locked @1.2vid I can get above 2400mhz on the NB.
I am not going to mess with the NBvid anymore or at least untill after another Bios. I dont need higher than 2000/2200nb ATm anyhow.
Its one thing to play with high volts Its another not knowing how high.:yepp:
NBvid @! 1.36vid in AOD 3.2..
I dont use AOD to overclcok .I only use the BIOS to OC on MOBO . So to me this sad that AOD can make it work ,but Gigabyte says theres no problem with NBvid,,its a problem:shrug:
http://i168.photobucket.com/albums/u...a/UD4Pvids.jpg
Note: when using AOD ,, It will set the NBvid and you can use Everest to see it change as well as other sensor software.
send this images to Gigabyte...
what is last BIOS for MA790X-UD4???
in czech gigabyte web:
http://www.gigabyte.cz/products/mb/b...0x-ud4_10.html F3F 19.3.2009
in global web:
http://www.gigabyte.com.tw/Support/M...ProductID=3002 F2A 10.3.2009 and F3F from 9.3.2009??? wtf?
I have a DDR2 version of this board (790X-UD4P) and I experienced one of the most strange things (not necessarily in a bad way) from a BIOS flash. After flashing to F5A, I saw this EC Firmware thing along with ACC and enabled it. And I saw a huge (relatively speaking) performance boost.
Before EC Firmware
http://altf3g.blu.livefilestore.com/...h/BeforeEC.jpg
After EC Firmware
http://altf3g.blu.livefilestore.com/...g4/AfterEC.jpg
Now, granted Everest cachemem varies each run, but I tried to pick shots that are averages. Memory Read went up by 500MB/s, and Memory Copy went up by 1000MB/s..! At this frequency (228HTT, 2.28GHz NB), Memory Read used to be always 200~300MB/s lower than L3 Read but now it equals or even exceed L3 Read.. Latency seems to be the same (ignore the latency in the 'before' shot because that latency is an outlier)
I also ran 3DMark06 CPU test and the score is consistently higher by 50~60 than before, @3.60GHz. Not a lot, but it's there.
What is AMD doing with ACC and this EC Firmware thing? Does anyone know what exactly it is?
I have with F3F bios nad 3600/2200MHz this:
http://img14.imageshack.us/img14/4240/cachemem3600.jpg
And 4GHz/2200 NB
http://img20.imageshack.us/img20/5443/cachemem4000.jpg
Interesting, so Based on Daveburts testing, and what AOD seems to be showing, the NBvid is not being adjusted via bios. Funny how AOD is showing the correct NBvid voltage, but not cpu voltage...or does your chip run 3.7 at 1.3 volts?>!?!! Also my AOD says my memory is at 1.7v when i ve set it to 2.0 in bios.Quote:
Originally Posted by gOtVoltage
I guess we ll see once we get the DMM readings :)
Ok, even more wackyness. When i try raising the voltages with AOD, ie raise nb volts to 1.36, apply... it comes up with the warning etc, then applies..but goes back to 1.2 on the scale...and everest shows the same thing.. no change?!