**Official DFI LanParty UT 790FX-M2R Review/Overclock/Guide Thread**

[keithlm]

Actually AWARD Flasher 8.83 is the recommended one from RGone and I agree with Robert on that one and its been hard road getting that figured out through trial and error on my end.

I was using the one that DFI has with their bios.... 8.90.

I went through 6 or 8 floppies trying to find one that would format... then I gave up and made a bootable USB stick... MUCH better. And much more stable than a floppy.

The last thing you want to hear while flashing is the floppy drive starting to do the "re-read" shuffle. Sadly while rebuilding the machine... the faceplate on the floppy came off... and I blew about 12 years worth of dust out of there. Then I realized this is the same floppy I've been using since 486 days... everything else has been changed multiple times.

BTW: When I hooked my drives up to the new MB/Phenom Vista updated all my drivers and everything worked very well. (I was ready to re-load... but never got around to it.) BUT going back to the old motherboard... it's all messed up and I HAD to reload. I think it has something to do with the ACPI stuff on the motherboards. (I'm guessing.)

[Brother Esau]

OK......I need some advise from users that Run Vista 32bit with 2X2GB Ram Kits please

I am New to owning Vista and I am also new to owning 2x2gb Kit of Ram and my question is this.....When using Vista what works well with Stability testing programs? Also what do I set in Prime95 for amount of Ram in MB to be used for Blend Custom?\\ I set 4000mb But Task Manager shows 2.16GB being utilized Is it just Vista or is it just me not being familiar with Vista or both?

So I need counsel on the Proper way to set up Prime95 Blend/Custom and HCI MemTest Windows using Phenom with 2x2gb sticks of Ram please

[aGeoM]

I can not set the CPU-FID if i set the cpu multi on the genius main page here.
Can you boot 3GHz at stock (1.3V) voltage?
And to nag you albit more, have you checked the max ref Ht you can get? Mine tops out at ~270MHz where the 9600BE did up to 285MHz. (on 3-07 bios).
Also I found the higher the ref HT the lower the possible max nb frequency can be.

Let's see..., for 4th time I'm gonna try reply to you guys.

Well, here it's the same, if I change CPU multi in Genius main page, no CPU-VID in CPU Feature page. What I was trying to say is, changing one at the time, didn't affect multi after booting Windows, because I was running at BIOS default, after CLR CMOS, loaded BIOS default, so CnQ was AUTO, and affected the multi setting in Windows, setting it at 12.5.
CnQ disable I can change CPU multi or CPU-VID to 15, and booted like that.

Yes, I can boot to Windows, 3000MHz default Voltage, in fact I was able to install Vista x86 too. But now (some hours ago) I'm facing IDLE lockups, only replaying here was enough to lockup 3 times in a row.

Maximum HTT was 275 at 11 multi, very unstable, enough to boot XP. 307 bios.

^^Looks like the CPU FID option has a major bug, it's not changing clock speeds at all [bit like Crystal CPUID for NB].

His 2800/3000 results are too slow. Wprime depends on CPU clock speed highly and his 3000MHz 32M is slower than my completely non-optimized, normal windows, AV/FW/FX/CPUZ running in background ran 32M at 2800MHz: http://img81.imageshack.us/img81/1269/2800wprimeav2.png

Here's my old 3000MHz run, its far quicker: http://img208.imageshack.us/img208/465/wp323gds4.png

And I've just ran my 2828 again [NB is stock, just changed it in CCPUID before to set on boot], its still fair bit quicker than his 3000: http://img214.imageshack.us/my.php?i...primehdid6.png

Which is not possible. His clocks are lower than 2800, his 3000 result looks like 2700. You have another DFI bug to sort out

But he has a very low stock Vcore, he has a lower TDP, higher binned model which should clock better than ours.

Hehehe... Tye that runs was 1st 9850 boot, BIOS on it's default as I explained above ( I guess ), multi changed in AOD, no reboot at all.

Maybe yes, system is running slow at 3000MHz general use, need couple secs to open this forum main page, not normal.

I got some EVE stress test at 1.37v, system is more stable, at least I'm still writing
Yesterday manage to boot 3100MHz at 1.34v, did CPU-Z Validation, lockup, end of story, tried 3200MHz, system boot, Windows load fail, I didn't wanted go over 1.36v, 1st CPU day., played some HT/NB/MEM at 3000MHz, IDLE instability. Went to bed, getting pissed off, grrr...
Gonna play a little more today.

Actually AWARD Flasher 8.83 is the recommended one from RGone and I agree with Robert on that one and its been hard road getting that figured out through trial and error on my end.

Have you or RGone test 8.87? 8.83 it's OK too, can be used /f command, 8.90
Have you more info about temps at 1.25v?
I haven't read all the new posts, just went straight to here for reply.

EDIT:

3DMark06: CPU=4081 AOD Yellow



EDIT 2:

3DMark06: CPU=4074 AOD RED



EDIT 3:

CPU=3000MHz
HTT=200
HT=2000MHz
NB=2400MHz
MEM=1066 CL5 5-5-16 2T Ganged

3DMark06: CPU=4160 AOD RED

wPrime :

[Brother Esau]

With my Tuniq Tower 120 Installed once again Temps are very normal

Currently I am @2.6ghz @1.23vcore Real

Prime95 Running Small FFT ambient is 80F Load I am 47c

[aGeoM]

Nice load temps.
Are you using the bios in your Sig?

[keithlm]

Anybody know if the DFI 790FX-M2R will work with a cheap $32.00 Sempron? The 790FX boards from other companies say they work with a Sempron... but DFI doesn't say anything. (But they don't say anything about using 1066 memory either.)

I'm considering ordering one of these as an "emergency" backup CPU. "Just in case". If I get my Motherboard back next week and the Phenom I have ends up being what my problem was... then I'd have to wait for that to RMA. In the meanwhile I could get a 2.0 Ghz Sempron for $32.00 and get it at about the same time that the motherboard comes back.

If the phenom works okay (which is my HOPE) then I would have wasted $32.00. On the OTHER hand if the problem IS the phenom... then the $32.00 would pay for itself immediately. (Since I wouldn't have to replace my old motherboard/cpu/memory yet again.)

So it would be kind of like "insurance". If I order it... then the laws of nature would say that it will be un-needed. :-)

PLUS as a bonus... I can use the Sempron with a small cheap motherboard and some cheap ram for building a multimedia server.

[Brother Esau]

Nice load temps.
Are you using the bios in your Sig?

Yes I am currently on 1/30/2008 Really don't understand enough or am unfamiliar with Phenom so Flashing Bios when you don't know whats normal and whats not is not a sensible Idea and not very bright at all even if I can reprogram the chip on my end I need a working Baseline on which to Judge other Bios on.

[Brother Esau]

Anybody know if the DFI 790FX-M2R will work with a cheap $32.00 Sempron? The 790FX boards from other companies say they work with a Sempron... but DFI doesn't say anything. (But they don't say anything about using 1066 memory either

.)

I'm considering ordering one of these as an "emergency" backup CPU. "Just in case". If I get my Motherboard back next week and the Phenom I have ends up being what my problem was... then I'd have to wait for that to RMA. In the meanwhile I could get a 2.0 Ghz Sempron for $32.00 and get it at about the same time that the motherboard comes back.

If the phenom works okay (which is my HOPE) then I would have wasted $32.00. On the OTHER hand if the problem IS the phenom... then the $32.00 would pay for itself immediately. (Since I wouldn't have to replace my old motherboard/cpu/memory yet again.)

So it would be kind of like "insurance". If I order it... then the laws of nature would say that it will be un-needed. :-)

PLUS as a bonus... I can use the Sempron with a small cheap motherboard and some cheap ram for building a multimedia server.

If its the same socket it (should) work. should being the operative word here

[aGeoM]

Yes I am currently on 1/30/2008 Really don't understand enough or am unfamiliar with Phenom so Flashing Bios when you don't know whats normal and whats not is not a sensible Idea and not very bright at all even if I can reprogram the chip on my end I need a working Baseline on which to Judge other Bios on.

Just ask, because was thinking to give 130 a try.

EDIT

CPU=3100MHz
HTT=200
HT=2000MHz
NB=2400MHz
MEM=1066MHz CL5 5-5-15 2T Ganged

3DMark06: Overall=16162 CPU=4258 AOD=Red





EDIT 2

wPrime



EVEREST



EDIT 3

CPU=3100MHz
HTT=200
HT=2000MHz
NB=2400MHz
MEM=1066MHz CL5 5-5-15 2T Unganged

3DMark06: Overall=16201 CPU=4278 AOD=Red

[Brother Esau]

Just ask, because was thinking to give 130 a try.

EDIT

CPU=3100MHz
HTT=200
HT=2000MHz
NB=2400MHz
MEM=1066MHz CL5 5-5-15 2T Ganged

3DMark06: Overall=16162 CPU=4258 AOD=Red





EDIT 2

wPrime



EVEREST

Appreciate the offer aGeoM So far I like this Bios with the Phenom but with my 6400+ BE it was just ok and not as good as 11/22/07 and the 3/12/08 Beta were but none the less good

With Phenom its a really nice Bios from what I see so far but also I have not really challenged the HT Bus deal either. Actually it appears that HT Clocking is kind of pointless on the Phenom seeing how you can manipulate the HT / CPU /NB all independently which is really, really nice by the way and jam on it that way But I suppose HT still finds its merit and worth in being able to incrementally increase and fine tune your clocks CPU/Ram/NB and what not But I really like the Phenom and it really rounds off the experience on this board quite nicely and Gaming is much more fluent even at stock compaired to My Q6600 @3.55ghz and 1186 on Ram now thats impressive Cant complain at all man not at all

[aGeoM]

Appreciate the offer aGeoM So far I like this Bios with the Phenom but with my 6400+ BE it was just ok and not as good as 11/22/07 and the 3/12/08 Beta were but none the less good

With Phenom its a really nice Bios from what I see so far but also I have not really challenged the HT Bus deal either. Actually it appears that HT Clocking is kind of pointless on the Phenom seeing how you can manipulate the HT / CPU /NB all independently which is really, really nice by the way and jam on it that way But I suppose HT still finds its merit and worth in being able to incrementally increase and fine tune your clocks CPU/Ram/NB and what not But I really like the Phenom and it really rounds off the experience on this board quite nicely and Gaming is much more fluent even at stock compaired to My Q6600 @3.55ghz and 1186 on Ram now thats impressive Cant complain at all man not at all

m8

You are working on 130, me and justapost on 307, and Tony on 331, 3 front lines.

Well... just finish for today, updated 3DMark06 ganged/unganged memory setting, had some Phenom love, I'm happy, I can go sleep now.

Didn't try Gaming on 9850, must install my favorite game ( Americas Army ) to play a little tomorrow, it's a good game for testing CPU too.

Cya guys, be well...

[Rammsteiner]

Brother, I hear that quite often from people both having an AMD and Intel setup. So guess that's a good and solid statement then

Anyway, I ordered a Phenom 9850BE, hopefully it will arrive today.

I asked this in the 790FX-DQ6 thread however no one seems to care or so.

Is it possible the IMC on my X2 is degrading or so? A few weeks back I had those strange crashes etc while my memory was allright. Turned out I had to downclock by 30Mhz and it was all good again.

Then this week I got them again. This time I thought it would be the memory (D9's), and I tested them with lower multi and they errored. However I wanted to be 100% sure and changed the DIMMs for one OCZ ReaperX PC6400 (two won't fit lol). Tested the CPU again at the 14x249 and it still BSOD'd. So it wasnt the D9's after all, since the IMC is linked with CPU speed and controlling the memory I think it might be the IMC dieing slowly?

Ill drop the CPU into my 790FX-DQ6 board today. It does have a broken 4th DIMM slot, but as long as I stay away from that it's ok. If I can get the CPU stable again on the DQ6 the problem must be at my M2R, however I really have the feeling it must be the CPU.

Any ideas? And even better, how to prevent it from happening next time.Temps never got over 50, and I only tried a very, very few times at 1.55V to check if I could OC further in the beginning.

Well, in this case I hope it's the CPU.I still have to RMA the DQ6 because of the 4th RAM slot. If it's the M2R, Ive to set everything over into the DQ6, RMA the M2R and when I get it back again change evrything over again and RMA DQ6.

[Brother Esau]

The only logical thing at this point is do as you say and drop it into another Board and further test to evaluate the problem and see if you can duplicate it in another set up. Could be the PSU as well that will also give you problems with RAM and make it error out been there done that one and was quite miffed by that situation until I finally came to the conclusion it was the PSU and sent the Siverstone STF06 in and got a PC-Power & Cooling and voila problem solved

The CPU? Doubtful but possible! Really depends on how hard you lean on your stuff with voltage applied but it does happen, as odd as a sighting of Big Foot at McDonalds But it is possible

Did you Test RAM with Memtest 2.0 yet?

[Rammsteiner]

Could be the PSU, but if that's true wouldn't the 'results' I got mean the PSU is degrading or something?

Ive a pretty good one (it's in my sig). But after all, not all PSU's are perfect. Silverstone is pretty good but well, sometimes a little design flaw wants some fun heh.

Ill try to get my Voltmeter and measure the Voltage under idle and load then. I dont hope it's the PSU as that does have quite an impact regarding rebuilding things.

Also my M2R makes a noise, but it's from around the CPU socket so eventually it's the CPU (which I don't think). Underload it barely makes noise, but idle it does make, and differing from the loading done at the moment the sound becomes different. I dont think this should be an issue. My old X800Pro did this, so did an older Epox sk754 board too. I heard this was just some component on a PCB not glued too well on it and starts to shake a fast, depending on the load.

Ill put a very last hand on the room cleansing before I start to test the things, though I hope Ill find the defect component very fast as I HATE to solve such things really.

One last remark, if it was my PSU, wouldn't it be more likely the PC would insta crash under load if the Voltage output becomes worse and worse? I mean, after all I actually used the same Voltages all the time, no matter what frequency. And from the first crash period, I only downlclocked 30Mhz, that wouldn't really lower Voltage/Ampere requirements by a significant number anyway (at same Voltage being applied it is).

[Brother Esau]

Could be the PSU, but if that's true wouldn't the 'results' I got mean the PSU is degrading or something?

Ive a pretty good one (it's in my sig). But after all, not all PSU's are perfect. Silverstone is pretty good but well, sometimes a little design flaw wants some fun heh.

Ill try to get my Voltmeter and measure the Voltage under idle and load then. I dont hope it's the PSU as that does have quite an impact regarding rebuilding things.

Also my M2R makes a noise, but it's from around the CPU socket so eventually it's the CPU (which I don't think). Underload it barely makes noise, but idle it does make, and differing from the loading done at the moment the sound becomes different. I dont think this should be an issue. My old X800Pro did this, so did an older Epox sk754 board too. I heard this was just some component on a PCB not glued too well on it and starts to shake a fast,
depending on the load.

Ill put a very last hand on the room cleansing before I start to test the things, though I hope Ill find the defect component very fast as I HATE to solve such things really.

One last remark, if it was my PSU, wouldn't it be more likely the PC would insta crash under load if the Voltage output becomes worse and worse? I mean, after all I actually used the same Voltages all the time, no matter what frequency. And from the first crash period, I only downlclocked 30Mhz, that wouldn't really lower Voltage/Ampere requirements by a significant number anyway (at same Voltage being applied it is).

That Silverstone model the STF06 was one of the Best PSU's available at the time and even Jonny Guru gave it a 9.5 and that was the last think I thought it would be ...but it was. Bad Power supply will definitely do that no power or not so good power in our systems means no worky worky properly!

A P.S.U going bad runs Hot & Cold just like women when its good its good when its bad well you know unpredictable and temperamental/

[Rammsteiner]

WTF!? So I just tested the PSU. Voltages were a little below spec, but AFAIK not super bad.

Anyway, I measure at idle. Pretty constant. Then load and it stays the same. Then after a few minutes I decide to test again and measure a lot less. 12V line almost hit 11V. And it's very very slowly getting less and less. At idle it rose a little again, but it rose a lot less then it actually dropped... wtf?

[EDIT]BIOS and Everest read higher Voltages though. I know software and BIOS could be wrong and a multimeter is pretty solid although this isnt an A-Branded multimeter

[EDIT #2]I just thought of this, the PSU could have been like this from the start. So no guideline to tell if it's PSU related yet although Ive to admit the measurements I did were quite strange though. I just put the CPU into the DQ6, 2x1GB and a 3870. Let it priming... THough I 'n miss the nice power butons on the mobo like the M2R has... Oh well.

I measured again with the multimeter and got the same values, although the DQ6 BIOS read more or less the same Voltage as the M2R... Right lol. However, as far as I know, the CPU is already priming longer on the original OC than on the DFI board with 11.12V measured on 12V line from the multimeter.

On the M2R I had of course a second 3870, 4x1GB total and another Raptor but well, Voltage is the same anyway. Ill let it prime longer of course but thus far the M2R looks like to be the bad boy.

Note to self: Look where you walk next time, almost stepped on top of my M2R+3870 on the ground lol.

[justapost]

Let's see..., for 4th time I'm gonna try reply to you guys.

Well, here it's the same, if I change CPU multi in Genius main page, no CPU-VID in CPU Feature page. What I was trying to say is, changing one at the time, didn't affect multi after booting Windows, because I was running at BIOS default, after CLR CMOS, loaded BIOS default, so CnQ was AUTO, and affected the multi setting in Windows, setting it at 12.5.
CnQ disable I can change CPU multi or CPU-VID to 15, and booted like that.

Aha ok.

Yes, I can boot to Windows, 3000MHz default Voltage, in fact I was able to install Vista x86 too. But now (some hours ago) I'm facing IDLE lockups, only replaying here was enough to lockup 3 times in a row.

Same here, tried 2.9/2.2GHZ at 1.4/1.35V ran everest stability test for an hour, tried a few applications before (including sandra benches) then i ran prime95 over night (65°C core temps in the morning).
Stopped prime95 and ran sandra latency test again and voila the system froze. Never had this random freezings with my 9600BE but that one did not like voltages.

Maximum HTT was 275 at 11 multi, very unstable, enough to boot XP. 307 bios.

Seems i just had luck with my 9600BE.

EDIT:
3070GHz/2456GHz at 1.525/1.5V system froze short after saving the screenshot.


Odin does odd power readings. With that PSU i had up to 480W DC power consumption. Replaced that PSU with an ZM850-HP and now i get a maximum of ~330W AC.

[nanohead]

Well, definitely puzzling behavior on my system as well. Got home last night, the 9850 BE was delivered yesterday. Removed the 5000BE, popped in the 9850, system booted right up. Rebooted, set the multi to 15 (x200) and it started no problem (@1.4v). Am water cooled, so temps never exceed 35 or so (assuming its somewhat accurate).

Ran Orthos for about an hour, no problem. Played Hellgate for about a hour, not problems.

Woke up this morning, booted, and BANG BSODs left and right. Did Safe mode, BSOD. Did normal boot. Hung. Then boosted NB-CPU volts (107% setting), started right up, went to start Everest, BANG crash. Rebooted, now running fine....

Using the 307 beta bios, had great luck with it.

[KTE]

Yeah, your scores are very low for your MHz. As if you've got big stability problems, that's why they occur - I had them on all Phenoms incl. 9850 if I push past 3070. Your CPU is benching 2800/2300 performance at 3000 settings. You might get more perf. with added voltage, that's lower than B3 perf for those clocks, so it should help. Same thing as mine with idle instability though - can get 3069 fully stable load but not stable idling. Vcore 1.2v boots 2.8G and 1.3v boots 3G into Windows, benches but will fail load/idle stability. 3100 1.4v runs EVEREST fine, does not run AOD stress test fine though. 3077 fails, 3069 is max load stable in everything not just one app, even in games, but still unstable for day in day out use.

The amount of random lockups I've had and the number of PMs/posts/emails I've lost.. only I can really know how depressing

[Oese]

just optimized a little...

and new pi1m

sorry for that much pictures..

[Brother Esau]

There ya go Oli Finally lets see what my old chip can really do

Give it to em Lefty

[Oese]

i need a 2gb set

I'll buy one when my old processor is sold...

[Brother Esau]

Wanna get on my FLEX 9200 before I sell them? I'll make you the offer first before I list them for sale.

[Oese]

oh boy you're scaring me.. hottest ram on earth.. how much including shipping??


edit: temps:

[aGeoM]

Hi guys.

Today, when booted Vista, instability returned, same config as last night, nothing was changed, I tough it was one of this no no days. Well if it don't want to go 3100, maybe it want 3200 , after some more volts, up where, down there, I manage boot 3200, check Everest, CPU-Z, etc, but when opening iexplorer, lookup again, never booted again, so loaded 3100 profile, and... were I am.

HT=2.2 in bios does not work, always 2.0 in windows, I had the option in AOD to go up to 2.2(x11) but sytem frozed.

Run some 3DMark06 bench, this time I had PCIe=124:

OVERALL=16256; CPU=4252; SM2.0=6280; SM3.0=8072