Asus M3A32-MVP Deluxe (790FX) *Pictures + Benchmarks + Overclocking*

[MrBallpoint]

I'm just bought myself this board a month ago, after flashing the BIOS to version 1202 now my experience with this board is proudly fine:-) But there is only one thing i can't understand and hope there is someone out here who can help me.

When i set de memory at his 1066 and do raising the htt with some crements it gaves me 565Mhz (1130). But when i do start Memtest the settings are weird and are totaly not correspondance to the settings i gave in the BIOS.



Is this a normal thing?

[Flasker]

That is normal. MemTest does not read the memory frequency or timings correctly in 1066 mode, still works for finding errors.

[msimax]

I'm just bought myself this board a month ago, after flashing the BIOS to version 1202 now my experience with this board is proudly fine:-) But there is only one thing i can't understand and hope there is someone out here who can help me.

When i set de memory at his 1066 and do raising the htt with some crements it gaves me 565Mhz (1130). But when i do start Memtest the settings are weird and are totaly not correspondance to the settings i gave in the BIOS.

Is this a normal thing?

hey what final oc would land you @ 1200mhz on the mem

[Phenomenalspeed]

Hi I flashed to it this morning as well.... nothing noticeable to report about it.... rally, no new settings or anything..

I havn't tired the NB-HT Speed setting yet to see if it will apply changes made to it... will try it later on...

Other than that it seems the only way to increase the HT speed is by increasing the Ref HTT.

I'm currently using the CPU Multi to OC, so maybe I'll lower the CPU and NB multi's and try increasing the Ref HTT to increase the HT Speed as well...

I'm thinking a HTT of 240Mhz with a NB Multi of x10 will give me a 2400Mhz HT speed and 2400Mhz NB Speed, and on a x12 CPU Multi should put it at 2.88Ghz for the CPU...

Though I'm not sure if the mobo will take to kindly to this...LOL...

Will post with details later on...

My motherboard/BIOS is very similar. No multlipler option in BIOS only frequency. It doesn't work though either.

[kazgirl]

My motherboard/BIOS is very similar. No multlipler option in BIOS only frequency. It doesn't work though either.

Yes its odd it seems that it will downclock, using the speeds as being representative of a multi decrease, so going from default x10 to x9 reduced my HT Speed to 1800Mhz, but trying to increase above x10 to x11 for a 2200Mhz HT Speed and even though it lets me change the setting and save and exit in BIOS it doesn't apply the new setting....

My NB speed is at 2400Mhz so that shouldn't be an issue, it should allow to increase up to and including the NB speed....

I've reported the problem to ASUS support, but as usual no reply from them....

BIOS support is a really..REALLY weak area of ASUS IMO.....!!!

[kazgirl]

Well I managed to work out the HT speed thing...

it seems that if you change the NB multi it then adds/removes options to the HT Speed, but it seems the max it will allow is 2.0Ghz in the HT Speed menu or so I've found so far...

I've currently got

260 x 11 = 2880Mhz
NB x9 = 2340 Mhz
HT 1.8Ghz (x9) = 2340 Mhz
RAM @ DDR2-800 = 1040 Mhz



Going to try 260 x 11.5 = 2990

Then if thats OK I might try 261 x 11.5 = 3001.5

261HTT would give:

CPU = 3001.5Mhz
RAM = 1044Mhz
NB Speed = 2349Mhz
HT Speed = 2349Mhz

I tried a 267HTT at lower multis but it failed to boot... not sure if this needed more HT Voltage...???

Was hoping to get 267 stable, to get RAM @ DDR2-800 @ 1066Mhz

267 x 11.5 = 3070Mhz CPU
267 x 9 = 2403 NB Speed
267 x 9 (1.8Ghz) = 2403Mhz HT Speed


I might try higher HT Voltage tomorrow.... and try the 267 HTT again..

[Phenomenalspeed]

Yes its odd it seems that it will downclock, using the speeds as being representative of a multi decrease, so going from default x10 to x9 reduced my HT Speed to 1800Mhz, but trying to increase above x10 to x11 for a 2200Mhz HT Speed and even though it lets me change the setting and save and exit in BIOS it doesn't apply the new setting....

My NB speed is at 2400Mhz so that shouldn't be an issue, it should allow to increase up to and including the NB speed....

I've reported the problem to ASUS support, but as usual no reply from them....

BIOS support is a really..REALLY weak area of ASUS IMO.....!!!

Well I managed to work out the HT speed thing...

it seems that if you change the NB multi it then adds/removes options to the HT Speed, but it seems the max it will allow is 2.0Ghz in the HT Speed menu or so I've found so far...

I've currently got

260 x 11 = 2880Mhz
NB x9 = 2340 Mhz
HT 1.8Ghz (x9) = 2340 Mhz
RAM @ DDR2-800 = 1040 Mhz

I think that possibly you misunderstand me slightly. I don't have this motherboard (See my signature). I was saying that my motherboard/BIOS has a very similar problem. Changes made to Hypertransport link frequency (no multiplier option same as your motherboard) in BIOS are not applied when you boot to windows.

So if I've understood you correctly (not sure if I have), changing the northbridge multiplier also indirectly changes the Hypertransport multiplier? But this removes/adds options from the BIOS setting which only allows a maximum of 2Ghz anyway.
So effectively you are using the HT reference clock and the northbridge multiplier to alter your Hypertransport multipler/frequency??

Digressing somewhat (off topic):


Your comments about ASUS support is very off putting. When I decided to get a 790FX+SB600 I was choosing between the top gigabyte (DQ6) and the top ASUS (deluxe Wifi). I will probably get a 790FX+SB750 motherboard sometime in the future and again was looking at these two manufacturers.

I only became aware of the Hypertranport problem after Gigabyte provided me with a beta BIOS fixing the northbridge multiplier. As I was able to increase northbridge multiplier I then tried to match the Hypertransport frequency and discovered the problem. Gigabyte are aware of it now and hopefully working on a fix. I was posting messages on Gigabyte's online technical service and I got a reply to each message a day later and an email notification. They matched my sytem setup and to test the northbridge multiplier. They reproduced the error. A day later they got back to be with a download link to a beta BIOS which fixes it. Even if the M3A79-T Deluxe proves to be an outstanding motherboard presumably ASUS's customer and BIOS support will remain unchanged. So if you do get problems you aren't likely to get much help from ASUS. Gigabyte's return mechandise authorisation process/customer support is certainly not flawless but what I have experienced so far is better than what I am hearing about ASUS.

I was reading the stuff posted by Kab and yourself on the AMD forums about the Voltage regulator module efficiency and instablity of this motherboard etc.
I wonder if such things have been improved with the M3A79-T Deluxe. Only time will tell I suppose.

By the way did you see my last few posts on the AMD Phenom forums? Username = Phenomenaloverclocker. I am a bit suprised with some of your results .I have the same case as you but your cooling is superior to mine.

[kazgirl]

I think that possibly you misunderstand me slightly. I don't have this motherboard (See my signature). I was saying that my motherboard/BIOS has a very similar problem. Changes made to Hypertransport link frequency (no multiplier option same as your motherboard) in BIOS are not applied when you boot to windows.

So if I've understood you correctly (not sure if I have), changing the northbridge multiplier also indirectly changes the Hypertransport multiplier? But this removes/adds options from the BIOS setting which only allows a maximum of 2Ghz anyway.
So effectively you are using the HT reference clock and the northbridge multiplier to alter your Hypertransport multipler/frequency??

No it sound s the same thing...

In my BIOS for the ASUS mobo...

Theres a NB multiplier settings with stock @ x10, (so x7,x8,x9,x10,x11.....x14 etc) Then there's a setting for HT Speed, which is in Ghz, with default @ 2.0Ghz and being the max setting it seems..

Now when I had an earlier BIOS when I had the NB multiplier at x10, the maximum setting avaialbe for the HT Speed was 2.0Ghz that was as high as it could go...

But when I increased the NB multi to say x12, the HT Speed would then provide addittional settings, of 2.2Ghz and 2.4Ghz, but if you chose one of those settings, they werren;t ápplied when booting into Windows..

With the latest BIOS 1202, its different again...

I now have the same options availalb:

NB Multi
HT Speed

But now when I change the NB multo to over the default x10, the HT Speed setting no longer provideds new settings to correspond to the higher NB Speed.... it stays as a max setting of 2.0Ghz.....

Now the speed settings in the HT Speed menu equate to being Multis, so:
1.6Ghz = x8
1.8Ghz = x9
2.0Ghz = x10

etc...

We know that when overclcoking the HT Speed shouldn;t be set higher than the NB Speed, i.e. (HT=/< NB) so this is obviosuly an ASUS way of making sure that users don't do this....

But it should be giving extra settings in the HT Speed menu when the higher NB speed is set, as the previous BIOS used to...

But at the moment all I can use is the HT Ref Clock to increase the HT Speed, rather than using a CPU Multiplier only to OC, which so far I think Phenom seems to give better performance from a multi OC rather than a HTT Clock OC.....

I've now changed to:

240 x 12 = 2880Mhz
240 x 10 = 2400Mhz NB Speed
240 x 10 (2.0 GHZ) = 2400Mhz HT Speed
RAM @ 960Mhz

Am testing this for stability at the moment.... before moving on to trying for 3.0Ghz and 2.4NB 2.4HT.... with 240 x 12.5...

If that proves unstable as 3.0Ghz has always been so far.... then I'll try for 2875Mhz CPU / 2.5Ghz NB / 2.5Ghz HT / 1000Mhz RAM

Digressing somewhat (off topic):

Your comments about ASUS support is very off putting. When I decided to get a 790FX+SB600 I was choosing between the top gigabyte (DQ6) and the top ASUS (deluxe Wifi). I will probably get a 790FX+SB750 motherboard sometime in the future and again was looking at these two manufacturers.

ASUS support has been awful for a long time now, I've never found it to be any good.... I havn't really got expereince of others to compare it to, other than what I;ve read on the forums of other peoples expereince of different manufactuers, but I think that it can easily be placed as one of the worst manufacturers for support.... or so it seems to me.....

I was reading the stuff posted by Kab and yourself on the AMD forums about the Voltage regulator module efficiency and instablity of this motherboard etc.
I wonder if such things have been improved with the M3A79-T Deluxe. Only time will tell I suppose.

By the way did you see my last few posts on the AMD Phenom forums? Username = Phenomenaloverclocker. I am a bit suprised with some of your results .I have the same case as you but your cooling is superior to mine.

Its a small world LOL.... yes I've seen the messages but been busy here and elsewhere and had some personal things to take care of as well... so havn't had a chance to reply as much as I would have liked....

I have been a bit surprised by some of the CPU Scores I've been getting at high clocks as well, compared to some.... I'm not sure what thats about unless its to do with the higher voltage requirements impeding CPU efficiency.... as I have to use very high voltage to get high clocks....

If it wern't for the WC I have, then I'd be stuck at much lower I think.... even for 2.8-2.9Ghz I'm having to use 1.4-1.475v and and for anything over 2.9Ghz needs more than 1.5v....

I'm really hoping that a SB750 mobo is going to help with that..... but will be looking forward to getting a Deneb to replace the 9850 later in the year...

I wass surprised by some of the VRM efficiency measurements that Kab had as well.... I didn't doubt them for a second, but I have no data myself to confirm it for my mobo..... but it would make sense.... I have to use a high CPU-VDDA 2.7 ATM to get better stability.

So I'm also hoping that M3A79-T Dlxe will have improved on things as well....

I'm still not 100% that I'm going to get that or a DFI mobo..... one of the 2..!!!

EDIT: tried the following settings for 3.0Ghz and it booted but again not stable, not even at 1.52v, so back down to 2.8Ghz ATM... going to have to wait for SB750 mobo it seems.

[tbone8ty]

heres a quick bench... i was playing with NB @ 2.4ghz




though the NB and SB heatsinks get super hot! 51C according to my scynthe temp prob. anybody had any luck with that NB/memory cooler? or other chipset coolers?

[Phenomenalspeed]

Its a small world LOL.... yes I've seen the messages but been busy here and elsewhere and had some personal things to take care of as well... so havn't had a chance to reply as much as I would have liked....

I have been a bit surprised by some of the CPU Scores I've been getting at high clocks as well, compared to some.... I'm not sure what thats about unless its to do with the higher voltage requirements impeding CPU efficiency.... as I have to use very high voltage to get high clocks....

If it wern't for the WC I have, then I'd be stuck at much lower I think.... even for 2.8-2.9Ghz I'm having to use 1.4-1.475v and and for anything over 2.9Ghz needs more than 1.5v....

I'm really hoping that a SB750 mobo is going to help with that..... but will be looking forward to getting a Deneb to replace the 9850 later in the year...

I wass surprised by some of the VRM efficiency measurements that Kab had as well.... I didn't doubt them for a second, but I have no data myself to confirm it for my mobo..... but it would make sense.... I have to use a high CPU-VDDA 2.7 ATM to get better stability.

So I'm also hoping that M3A79-T Dlxe will have improved on things as well....

I'm still not 100% that I'm going to get that or a DFI mobo..... one of the 2..!!!

If you have had long term bad experience with ASUS support why do you stay with them??

[Hemi]

Kazgirl, I know how you feel with your board. The best i could get is 3.0 prime stable using vista64 with air cooling. Any higher and cores 2 and 3 would fail prime. I just used the mutli to overclock. The chipset voltage including southbridge made a difference on the stability. I just picked up the foxconn a79a-s. Right now I am running 3.2 and prime 95 has been running about 45 minutes stable. Temp is up to 65c, need water cooling. I know you will love the 750 when you get one and it is very stable. Only had board about 24 hr"s. Get one as soon as you can.

[HondaGuy]

Just doing some testing on this 9850 so far OC @ 3.5
Seems to be stable

[Nedjo]

HondaGuy you'll need to edit sig

P.S
where can I dload this new BIOS?

[kazgirl]

Just doing some testing on this 9850 so far OC @ 3.5
Seems to be stable

Awesome overclock there, very...very lucky Phenom chip you have there, Phenoms that will OC as well as that on air and on SB600 are very..VERY scarce....

I wish my Phenom would get anywhere near that.... my temps aren't an issue as I;m on water, and Power is fine as well.... but no matter what the max it will do stable is 2.9Ghz.... waiting on a SB750 mobo now, hopefully the next 2 weeks will see them available here in the UK..

HondaGuy you'll need to edit sig

P.S
where can I dload this new BIOS?

Last I checked it was available on the ASUS downloads page.... though I think there's a link in another thread entitled M3A32-MVP New BIOS here as well......

[HondaGuy]

Here is Bios 1203. On ASUS site sometimes it there, other times its not. Here is the BIOS that I'm using right now

http://rapidshare.de/files/40345157/1203.zip.html

[radaB]

Anyone get better OCing with this bios compared to previous? Anyone know a changelog for it?

[kazgirl]

I didn't get higher OC but I did get same OC as previous but stable at lower voltage than needed before for some reason.

The changelog mentions fixing problems with the C1E enhanced power state, which is similar to Col 'n' Quiet, but I'm not 100% sure what it does whether it maybe dynamically changes chipset speed or similar, I think someone else mentioned that enabling the C1E as well as C'n'Q resulted in lower power consumption...

I havn't got C1E enabled myself, or C'n'Q either so I can't comment on it myself, though I've been using the 1203 BIOS for a few days now with no problems...

[Serj_Over]

Hi,
Does someone have a problem with board, when PC turns off for few seconds after applying some changes in BIOS? I'm using Athlon 5600+
Is there any fix for this problem? BIOS update wasn't usefull
Thanks!

[justinl61111]

Hi,
Does someone have a problem with board, when PC turns off for few seconds after applying some changes in BIOS? I'm using Athlon 5600+
Is there any fix for this problem? BIOS update wasn't usefull
Thanks!

It is supposed to do that.
Some BIOS changes don't take unless the board is reset like that

[Oliverda]

up

[Nedjo]

Hi,
Does someone have a problem with board, when PC turns off for few seconds after applying some changes in BIOS? I'm using Athlon 5600+
Is there any fix for this problem? BIOS update wasn't usefull
Thanks!

It is supposed to do that.
Some BIOS changes don't take unless the board is reset like that

YES! that's ASUS

[HondaGuy]

Kazgirl: Here are my setting dealing with NB and HT in BIOS,, I have mine set at NB 2600, HT@2600 CPU 3.3

[Oliverda]

Kazgirl: Here are my setting dealing with NB and HT in BIOS,, I have mine set at NB 2600, HT@2600 CPU 3.3

Why did you raise the HT Voltage, Core/PCIe Voltage and the NB PCIE PLL voltage? Do they have any effect to OC?

I have M3A79-T board and the HT multi switch does not have effect when I try to set over 10.00x (2000MHz).

[kazgirl]

Kazgirl: Here are my setting dealing with NB and HT in BIOS,, I have mine set at NB 2600, HT@2600 CPU 3.3

Hi Honday Guy

Thanks...!!

I'll give those settings a try next time I re-boot, and let you know how they work out....

I'm not optimistic though...anytime I've tried for a NB/HT speed over 2.4Ghz I failed to POST, I'll tryit again with the HT Iso Flow enabled to see if it helps...

Thanks again

PS: I've eidted my sig to show the correct BIOS now as well.... I updated to v.1203 a while ago, forgot to change my sig though..

[HondaGuy]

Why did you raise the HT Voltage, Core/PCIe Voltage and the NB PCIE PLL voltage? Do they have any effect to OC?

I have M3A79-T board and the HT multi switch does not have effect when I try to set over 10.00x (2000MHz).

When doing Cinebench there was an increase...Time was1:07 score was around 13196
Sometimes when I booted up it didnt take in affect untill I changed it in AOD



Here is NB@2600 HT@2000 CPU@3.3