**Unified** DFI LanParty UT 790FX-M2R-M2RS / Phenom /Data Base Thread**

[Brother Esau]

Give what I said a go . (Don't Cheat ) I will be watching

[Oese]

@ campbell: up and preparing for dmm measurements

@ pulsar: I think one can use any previous bios but then the pII will not be recognized correctly. If you have a spare cpu (x2 for example) possible to try. I heard from m2rs users that they use older bios and say its more stable with pII, only drawback you dont read pII at the boot sequence xD dunno...

[Brother Esau]

I am here Olli

[Oese]

damn campbell i startet with 1,3625v instead of 1,35625v and just flashed from 01/16 to 26/12

will take a bit longer now xD i have to do 01/16 again or perhaps proceed to 01/26 from 12/21 instantly and compare those two..

[Brother Esau]

Yes start at 1.35625 reason being is that it registers as 1.34v in Bios and also CPUZ & OCCT curious to know what it truly is on DMM

[Oese]

campbell i can give you a first result: with my system, 1,35625v shows 1,36v in bios with 12/26 not 1,34v and cpu-z as well.

and another thing: comparing 1,3625v,

26/12 is DMM idle 1,372v/load 1,384v
01/16 is DMM idle 1,371v/load 1,382v

so not that big difference!!! it seems like vcore is not the problem here at least if you do not use % table...

i'll test further...

p.s.: differences in voltage between my sys and yours could possibly stem from different 3,3v/5v/12v power supply lines? because the way the bios reads them voltages out should be the same right? or its tolerances in the electronic parts to read out and control voltages...

couldnt you get a DMM today (its saturday overseas as well right)? these are not really expensive and VERY helpful. So we could compare better.

next results: 1,36875v:

12/26: idle 1,378/load 1,390
01/16: idle 1,378/load 1,390

i have to test nb voltage as well...

1,375v:

12/26: idle 1,385v/load 1,396
01/16: idle 1,385v/load 1,396

i think here it becomes clear that vcore isnt the problem.

i flash to 01/16 now again to compare 1,35625v and nb be back in a minute :/

[Brother Esau]

campbell i can give you a first result: with my system, 1,35625v shows 1,36v in bios with 12/26 not 1,34v and cpu-z as well.

and another thing: comparing 1,3625v,

26/12 is DMM idle 1,372v/load 1,384v
01/16 is DMM idle 1,371v/load 1,382v

so not that big difference!!! it seems like vcore is not the problem here at least if you do not use % table...

i'll test further...

p.s.: differences in voltage between my sys and yours could possibly stem from different 3,3v/5v/12v power supply lines? because the way the bios reads them voltages out should be the same right? or its tolerances in the electronic parts to read out and control voltages...

Don't know would need DMM and 2 different P.S.U to confirm yes or no

So from what I am seeing with DMM Reading is that Voltage set in Bios = More @ Actual Value so in short Bios set is over volting @ idle and more so at load...................Value set in Bios should be Value Read on DMM plain and simple. No point in anything else or else its just a false sense of security thinking you are using less then actual.


Man I need to get another DMM but don't have the $160.00 to Buy the FLUKE I want and I refuse to buy cheap inaccurate tools

[Oese]

thats right with that fluke thing. i have a rather good dmm but cannot guarantee 100% accuracy in the mV range though... but at least relative comparisons should be possible with 100% acc and there it shows nothing has changed between 12/26 and 01/16 vcore wise if you dont use %.

i will check last settings and nb now... then test 01/21 even if its not as good as 26/12 possibly... man can it be :shake:

vdimm for your information is same with both bios versions as well. i get 2,044 reading with bios set at 2,01v with both 26/12 and 01/16...

[Oese]

lol i was there right in time i think

ps: i must first install skype perhaps lets do it after you slept a bit??? must be around 3-4 in the morning over at your side

i planned to lap my cpu and do some cold weather oc to break some forumdeluxx amd pi rekords after the bios thing. and i want to test wether 12/26 is the better bios over here as well... lots to do today xD

p.s.s: that program sounds nice

OK last comparison:

1,35625v

12/26: idle 1,364/load 1,375
01/16: idle 1,365/load 1,377
01/21: idle 1,365/load 1,376

so here 01/16 is even higher in voltage!

comparing nb its the following:

12/26: idle 1,178/load 1,177
01/16: idle 1,178/load 1,175
01/21: idle 1.178/load 1,175

so PERHAPS it could be the nb voltage that is lower then with 12/26. but i cannot believe it -3mV would be the cause...

whatever i'll testing 01/21 now...

[Brother Esau]

mmmmmm........Interesting to say the least Don't know what to make of your findings. I thought for sure DMM would show stark difference between both Bios wonder what the deal is?

Both 1/16/2009 and 1/21/2009 "Beta" show slump on 5v Rail in OCCT Graph I even Plugged in my PC-Power & Cooling 750w to see if the SilverStone was the issue and had same effect with 750w P.S.U as well

I have another Phenom II 940 BE coming and a LanParty Jr 790GX-M2RS coming in a couple of days. The 790GX is for a HTPC Build I am doing thats being sold the other 940 BE is to see which is the better clucker and compare to first on both 780FX-M2R & 790GX I wish D.F.I would send me a Board of the new 790FXB - 790GX series so that I can test Bios on them as well. I have no interest in giving up my 790FX-M2R just a better board minus the SB600 of course

[Oese]

yeah, actually with my balcony oc session today i would say the 01/21 is indeed the better bios overall compared to 12/26. i had no luck with it at all.

but i'm not pleased with my results whatever, will have to look under the cooler again.

btw thats how it looks like over here atm:

[bedlamite]

My new mobo came
http://img88.imageshack.us/my.php?image=tempyd1.png
CPU is temporary till I'll get some 945.

[Brother Esau]

yeah, actually with my balcony oc session today i would say the 01/21 is indeed the better bios overall compared to 12/26. i had no luck with it at all.

but i'm not pleased with my results whatever, will have to look under the cooler again.

btw thats how it looks like over here atm:

I like your place olli So you say No Luck with Which Beta 12/26/2008 or 1/21/2009 beta?

[Oese]

hehe thx i'd say 12/26 is less good. but thats because i got 4,14ghz with 01/21 a few times, with 26/12 i was at 4,05ghz max only..

26/12 is overvolting a lot. 01/21 gives 1,61 or something at 1,6v set in bios, 26/12 gives 1,634v

but thats apparently only the case with those higher voltages as we have seen... i did not measure any differences in that range we have tested today morning..

to be honest, i'm a bit confused atm. my chip does not do that well overall, and sometimes i measure these high voltages... i feel like something is still wrong, but i dunno if its the bios. perhaps its the sb600 really, so that i'm not able to reach 4,2ghz although the chip is chilled at 20°C... some ppl say acc is implemented in the cpus since pII. does the sb have to support it as well? will this chip go beyond 4,2ghz with an other board? i dunno. i believed, the m2r will be one of the best boards for pII around as pII doesnt need acc at all...

[pershoot]

hehe thx i'd say 12/26 is less good. but thats because i got 4,14ghz with 01/21 a few times, with 26/12 i was at 4,05ghz max only..

26/12 is overvolting a lot. 01/21 gives 1,61 or something at 1,6v set in bios, 26/12 gives 1,634v

but thats apparently only the case with those higher voltages as we have seen... i did not measure any differences in that range we have tested today morning..

to be honest, i'm a bit confused atm. my chip does not do that well overall, and sometimes i measure these high voltages... i feel like something is still wrong, but i dunno if its the bios. perhaps its the sb600 really, so that i'm not able to reach 4,2ghz although the chip is chilled at 20°C... some ppl say acc is implemented in the cpus since pII. does the sb have to support it as well? will this chip go beyond 4,2ghz with an other board? i dunno. i believed, the m2r will be one of the best boards for pII around as pII doesnt need acc at all...

get your hands on an fxb-m2rsh and try it out

[Brother Esau]

lol i was there right in time i think

ps: i must first install skype perhaps lets do it after you slept a bit??? must be around 3-4 in the morning over at your side

i planned to lap my cpu and do some cold weather oc to break some forumdeluxx amd pi rekords after the bios thing. and i want to test wether 12/26 is the better bios over here as well... lots to do today xD

p.s.s: that program sounds nice

OK last comparison:

1,35625v

12/26: idle 1,364/load 1,375
01/16: idle 1,365/load 1,377
01/21: idle 1,365/load 1,376

so here 01/16 is even higher in voltage!

comparing nb its the following:

12/26: idle 1,178/load 1,177
01/16: idle 1,178/load 1,175
01/21: idle 1.178/load 1,175

so PERHAPS it could be the nb voltage that is lower then with 12/26. but i cannot believe it -3mV would be the cause...

whatever i'll testing 01/21 now...

OK ...Back on 1/21/2009 "BETA" now from studying what DMM Reads @ 1.35625v on 12/26/2008 - 1/16/2009 - 1/21/2009 we can see that at that particular setting we can see out of all 3 different Bios that 12/26/2008 runs lower voltage @Idle and Load so what we see in actual fact is 12v Rail DMM Reading..........But when we look closer in theory or in actual fact 12/26/2008 "Beta" should be the most likely Candidate out of all 3 Bios to Fail OCCT CPU High Priority So ask yourself whats the cause of Failure on 1/16/2009 & 1/21/2009 I would say the problem does indeed lie in the Bios Voltage Table but it resides in Voltage allocation to 3.3V & 5.5V (Excessive Ripple) to those 3.3v & 5v Rails that is causing premature Failure in OCCT Test by comparing the Graphs generated by OCCT with all 3 different Bios 12/26/2008 .... Flatlines across all Voltage Rails.......1/16/2009 has Excessive (Ripple) and Drrop under Load on 3,3v Rail .......1/21/2009 has Excessive (Ripple) and droop under load on 5.5v Rail.

Now without a Oscilloscope you cannot Put this to affect and confirm what I am saying. DMM is handy but far from enough to troubleshoot this situation.


I also Notice present on all Bios that some areas of Voltage set is accurate where others are inconsistent and Fluctuate as you move up the Voltage Adjustment in Bios.

[Oese]

that could indeed be a factor... could you post occt graphs? i'd check mine for that...

here it is. 12/26:



01/16:



and again 01/16, but the 940:



apparently, for the 920, 01/16 indeed shows more ripple. But not for the 940...

[Brother Esau]

Reserved>>>>>>>>>

[Rammsteiner]

to be honest, i'm a bit confused atm. my chip does not do that well overall, and sometimes i measure these high voltages... i feel like something is still wrong, but i dunno if its the bios. perhaps its the sb600 really, so that i'm not able to reach 4,2ghz although the chip is chilled at 20°C... some ppl say acc is implemented in the cpus since pII. does the sb have to support it as well? will this chip go beyond 4,2ghz with an other board? i dunno. i believed, the m2r will be one of the best boards for pII around as pII doesnt need acc at all...

Hmmm, I hope not it's the SB600 cause Im planning to move back to the M2R real soon

Did you try different HTT and multiplier setups to get higher speeds? Also may I remind you, if you didnt do that yet, you'll have to run your NB at a minimum speed of 50% of your CPU speed?

I couldnt get over 3.6Ghz initially on this board after I remembered vaguely that either the HT or NB speed should be => 50% CPU speed. As soon as I ramped the NB up I could clock further.

Also Ive ACC disabled currently, didnt use it while enabled though, but it doesnt seem like I miss anything Some people use 780a chipsets and have some nice speeds as well, although then again there were some nice 9850/9950BE batches which hardly suffered from the SB600.

So if someone could find out whether SB600 is still limiting PhII OC's I'd really like to know it ASAP. I really want the M2R back in the system cause of the digital PWM's (although they make noise), but that seems a lot more solid than those retarded DK boards

[Brother Esau]

Its not the SB600 its the Bios they are screwy. But as stated I am getting a DFI JR 790GX and another 940 BE and will compare both chips in both boards and if the SB750 adds anything to this equation I'll eat my hat!

[Oese]

y i dont think its the nb either.. heard of lots of people that had problems to clock their pII on sb750 as well. its simply a matter of good or bad chip (or bios..)

[bldegle2]

After some time with the 940's and the SB600 from DFI I am of the feeling that it is time to move on to another board.....the board and/or the bios is suspect, this is not to say it is a bad board, worked great with a 6400+ but I know in my heart it is limiting the PhenomII....

Moving on, prolly FoxConn w/sb750.....it is really the only way to find out for sure, using all the same components except for a new MB.....

I haven't even tried the 1/21 yet.....I am quickly losing interest...

Laterzzzz................

[Oese]

yeah, the foxconn are good and cheap now, but the bios support will cease soon...

whatever let us know your results when comparing those 2 boards.

meanwhile, i came to suspect the bios as well, but in this theres no difference between 12/26 and 01/21:

both of my processors wont boot into xp with 4ghz, regardless if i use multipliers or htt oc.

i dont believe this is because they are both crappy ocers, then it would be an extreme weird case.

one of them runs 3,64ghz in vista64 prime and occt high priority with 1,508v, so not the very best oc indeed. but it scales well with voltage up to 1,6v to get into xp at 3,94ghz.

the other one runs prime @ 3,5ghz @ stock and @ 3,8ghz in vista64 (under x64 environment, prime is more demanding then occt i have found out, prolly because occt is not 64-bitted) with only 1,472v when using multiplier oc, but wont boot at above 3,94ghz into xp either. That is with 1,5v, but with above 1,6v as well.

tried different multi/ht/nb setting, no luck. With cold, it did 4,14ghz one time but no more... i did expect to hit 4,2ghz easily, as other similarly cooled setups (water or even chilled water) are able to reach 4,4ghz with a similar scaling processor... but no way.

i really think its the bios.. that is because other strange things happen, like hanging on boot with code 75 or CF needing manual shutdown and switching of power supply etc...

That is with 01/21 bios, but with 01/16 and 12/26 its no better...

perhaps i should try out pre-pII bios if that behaves differently...

[bldegle2]

Every really good clock on the PhenomII is on other boards, yes, I have seen a screenie or three of someone hitting over 4 gig on a SB600, but the followup comments posted with those threads usually indicate stable operations @ about 3.7/3.8 gig, which is what I am running as are many out there in the wild....

despite what one may think, there is only so much you can do with a bios, the board was built and designed before the PhenomII was more than a schematic on paper.......every MB intineration has its quirks.....

is the SB600 trully limiting clockage. my money is on 'yes'....

the one thing that still pesters me is the PWM temps, most of the time they are totally under control, but there is something that happens every once and awhile and it shoots to 53*c at idle and then takes off even higher when benching, and the only way to get it back to the high 30*c is to shut down, unplug and discharge the PS, leave it unplugged, and come back in 20 minutes or so and it will be good to go again, when in its 'good' mode it tops out @51*c full load....something is happening in the background, it mystifies me....it got a bit better with the newer bios's, but yesterday morning it reared up again.....

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

"yeah, the foxconn are good and cheap now, but the bios support will cease soon..."

what does this mean, are they going out of business too?

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>




laterzzzz..................

[Brother Esau]

well the reason not many people bought this board is because of the $260.00 price tag on it and lets face it other boards at that time were and still are having issues with Phenom /Bios so its not isolated to this board.

I am frustrated with the whole bios issue too considering the amount of time I have put into it Over year now. IO really think that the Bios issue will be corrected across all makers but once again I guess it comes down to money and putting your resources where most profitable and for all board makers that would be the Intel division of the company.


So now and just recently is AMD regaining ground and [popularity once more which is what the past month maybe? So you honestly can't expect miracles over night.


Foxconn may be a good board but you will see as Oese said Bios support will not exist soon


Also from what I see 3.5ghz =- 3.6ghz will be the average stable overclock not 4.0ghz that would be really nice but you would have to be really lucky to have fate smil on you like that.