M3A overclocking

[justapost]

Thanks for your input justapost .
What's the max. stable clock you achieved with 9850 with 4 cores enabled and with what Vcore/FIDs?
Also,was it stable in everyday usage,ie. not crashing in idle state?

Only tried different multis with the 9850e. The limit on the board is the single 4pin 12V connector.
At 2.8/2.2GHz 1.325/1.3V (cpu/nb) it pulled ~12A.
At 3.0/2.4GHz 1.525/1.5V it pulled ~19A, that exceeded the 18A limit of my PSU and the system powered of.

2.8/2.2GHz seems to be stable at above voltages. Have not tested it that much on the M3A but it worked fine with slightly lower voltages in the Sapphire 790FX mobo.

[informal]

Ok,thanks .
So the limit was actually your PSU?Since i doubt the 4pin 12V connector is limited to 19Amps.
BTW what PSU you have?

[justapost]

Ok,thanks .
So the limit was actually your PSU?Since i doubt the 4pin 12V connector is limited to 19Amps.
BTW what PSU you have?

I use the Gigabyte ODIN 550 atm. It comes with monitoring software.
http://www.gigabyte.de/Products/Powe...%20GT_550W
Also I have an ZM-850 here which is also limited to 18A on 12V1 but allows higher amps for all 12V connections together.
Do you know a PSU which can handle more than 18A on that 4pin rail?

[informal]

I think Antec TruePower 2.0 550w has 2 12V rails with 19 Amps max,at least for the sub 100$ price region.
But you are correct in saying that PSUs with higher than 18A amperage are not so common.

[justapost]

I think Antec TruePower 2.0 550w has 2 12V rails with 19 Amps max,at least for the sub 100$ price region.
But you are correct in saying that PSUs with higher than 18A amperage are not so common.

Thanks,

Beside that mosfet temps get realy high here atm. I placed a temp probe from the PSU on a mosfet.

80°
Testing 3150MHz (300x10,5) here with the 4850e. CPU temp is no issue ~60°C.

EDIT: ref HT 400MHz did not work with higher CPU multis than 5.5.

[informal]

Ouch @80°C !
Place some power fan if you have a spare one,over the mosfet section.I did this and mosfets are around 55-60° Cels. now.
From reading latest update at AnandTech(about the mosfet problems on 780G based mobos),it seems most of them come close to burning up @ around 120-140°!
I would say anything above 100° is danger zone.
http://www.anandtech.com/weblog/showpost.aspx?i=427

While a board’s PWM/MOSFET count generally indicates its capability to handle a particular load rating, it does not always indicate its ability to properly regulate, correct, and deliver clean power. That is why a board with a properly designed four-phase system can offer improved quality/performance over a poorly designed five- or six-phase system. All of our boards in the roundup utilize a decent three- or four-phase PWM design, with the exception of the ASRock board that features a five-phase design. Does this mean that the ASRock board will handle the 9850BE without a problem? Not necessarily, as there is more to our story than PWM design.

The critical aspect of running a 9850BE at stock speeds on the current 780G boards centers on the cooling of the PWM circuitry. Each and every manufacturer along with AMD agreed that cooling the MOSFETS properly was critical to the successful operation of the board at stock or overclocked speeds with the 9850BE - and to some degree, the 6400+ X2. The lack of space around the CPU area in the uATX design results in a crowding of components leading to additional thermal loads. This type of thermal loading is not prevalent in most ATX designs utilizing similar PWM components or the space that an additional/improved phase design (with associated components) affords.

We have measured MOSFET temperatures ranging from 128C~146C utilizing the stock CPU cooler on the 9850BE in our Silverstone SG03 case (temps were about the same in our Thermaltake and Zalman cases). While cooling of the PWM/MOSFET area is critical, most of the manufacturers admitted that improved component choices and electrical layouts similar to the 790FX boards is also required to ensure a robust design for users who are on the more enthusiastic side. However, this means additional costs that would place the board out of its intended market, a market where probably 98% of users would not worry about running a 125W TDP processor. For those who demand this extra level of performance, we expect to see several 780G ATX designs in May offering improved PWM designs.

BTW,very nice AthlonX2 chip you have there

[justapost]

I was courious if the airflow from the new cooler (see sig) is sufficient. Seems not. Added a fan and the probe reads ~33°C now. That anandtech article seems to be not overblown.

In comparison to the other X2's i tested (5000BE,BE2400) this one needs ~0.15V less for 2.5GHz. In terms of overclocking they all require 1.475V for 3.125GHz.

I tested an Intel 8400 (3GHz) with an p5ke-wifi mobo with the same peripherials (only difference I used two hd's in raid0). That required ~60W DC idling and ~105W DC under prime95 load, of course at 3GHz.

Gonna try 3GHz (300x10) with that X2 now, to see how they compare power consumption wise clock for clock.

My current 3.125GHz setup requires 58W DC in idle and ~145W DC under load.

EDIT:

CPU 3GHz (300x10)
1.300V restart during windows startup
1.325V prime95 failed on core 2 after ~4minutes
1.350V seems to do it

idle: 58W
load: 129W

I added another 500GB hd and an amd stock cooler for the mosfets so beside the mobo and the cpu the setups are identical.

For a rough energy efficiency comparison i ran CB10 multi cpu test and noted the time used and the power consumption.

4850e(3GH): 190s * 118W = 22420 J
i8400(3GH): 139s * 88W = 12232 J

At work where most of tasks can be done in more or less idle mode I still prefere the amd x2 cpu's and the power consumption of this chips at stock is amazing.

A more office related test I run on all setups is opening an huge openoffice spreadsheet. Here both setup require ~57s, so they are on par clock for clock here.
I found the same comparing an qx6850 with the 9600BE.

[justapost]

I was curious what voltage this cpu needs for 2GHz.

It primed fine for ~8hrs at 1V.
Currently i test 0.95V. Here the system consumes 36W DC idle and 58W DC under prime95 inplace load.
P-Tuners shows 0A in idle and 1.1A max under load at those settings for 12V1.
I removed the second HD and the two additional fans, to get comparable results with my previous testings add ~12W to above results.
I terms of power consumption a really amazing chip.

[Extelleron]

I use the Gigabyte ODIN 550 atm. It comes with monitoring software.
http://www.gigabyte.de/Products/Powe...%20GT_550W
Also I have an ZM-850 here which is also limited to 18A on 12V1 but allows higher amps for all 12V connections together.
Do you know a PSU which can handle more than 18A on that 4pin rail?

If I have a PSU w/ one big rail (Corsair HX 520) does that mean I would not run into a problem w/ power even at the voltage you are talking about?

Note that the HX 520 officially has 3 12V rails but according to johnnyguru's review it has 2 at most and no over current protection for those rails.

For people w/ M3A + Phenom, is there anything good or bad you have to say about this board? I'm thinking that I am going to go with this board for my Phenom 9500. Doesn't look perfect but it looks like one of the best 770 boards and I'm not looking to spend >$150 on a 790FX board.

[justapost]

If I have a PSU w/ one big rail (Corsair HX 520) does that mean I would not run into a problem w/ power even at the voltage you are talking about?

Note that the HX 520 officially has 3 12V rails but according to johnnyguru's review it has 2 at most and no over current protection for those rails.

I'm not an expert in this area, KTE has so glad to give me an short crashcourse.
It's simple math i guess.
That corsair PSU has two 4pin rails (1 8pin if you prefer), on the M3A there is only a 4pin connector so you can only use one of those two.
I checked the specs and the three 12V rails all support 18A but all together only 40A (GBT Odin support 41A for all three).
With an board with just one 4pin connector you are in the same situation that I am with the M3A.
Assuming you run a phenom in a board with an 8pin connector you would not be limited by 18A but 36A. But then you would have only 4A left for the third 12V rail (pcie).
1.5V is too much for the phenom so ~20A is the max consumption you can expect for 12V1+12V2.
So in the worst case you whould have ~20A left for the 12V3. (240W).

[Extelleron]

I'm not an expert in this area, KTE has so glad to give me an short crashcourse.
It's simple math i guess.
That corsair PSU has two 4pin rails (1 8pin if you prefer), on the M3A there is only a 4pin connector so you can only use one of those two.
I checked the specs and the three 12V rails all support 18A but all together only 40A (GBT Odin support 41A for all three).
With an board with just one 4pin connector you are in the same situation that I am with the M3A.
Assuming you run a phenom in a board with an 8pin connector you would not be limited by 18A but 36A. But then you would have only 4A left for the third 12V rail (pcie).
1.5V is too much for the phenom so ~20A is the max consumption you can expect for 12V1+12V2.
So in the worst case you whould have ~20A left for the 12V3. (240W).

Well it does have 3 rails officially, but according to johnnyguru it appears to have either 2 rails w/o any overcurrent protection (so they could run >18A) or one big rail: http://www.jonnyguru.com/modules.php...=Story&reid=18

This demonstrates that if the two rails are in fact separate, there is no OCP (over current protection) on each rail. Outside of a few traces zig zagging across PCB, I couldn't find how even 12V1 and 12V2 are separate, but I'm going to give Seasonic (the OEM for the Corsair units) the benefit of the doubt and say that we seem to have two 12V rails here, neither with any kind of "limit" on them.

Granted I don't really know what I'm talking about here. Power supplies are not my area of expertise.

[justapost]

Well it does have 3 rails officially, but according to johnnyguru it appears to have either 2 rails w/o any overcurrent protection (so they could run >18A) or one big rail: http://www.jonnyguru.com/modules.php...=Story&reid=18

This demonstrates that if the two rails are in fact separate, there is no OCP (over current protection) on each rail. Outside of a few traces zig zagging across PCB, I couldn't find how even 12V1 and 12V2 are separate, but I'm going to give Seasonic (the OEM for the Corsair units) the benefit of the doubt and say that we seem to have two 12V rails here, neither with any kind of "limit" on them.

Granted I don't really know what I'm talking about here. Power supplies are not my area of expertise.

That clarifies things. However I can not follow why JonnyGuru assumes there is no overprotection at all. Looking at this image i'd expect 12V1+12V2+12V3 are overprotected together at 40A max.
Seems you do not have a 18A limit with the PSU. Nice.
It should be a great PSU for the MSI 790FX Platinum which has only one 4pin connector but better PWM than the M3A.
You'll run into mosfet temp issues which such loads on the M3A no doubt.

Another note in the Sapphire/DFI mobo with an 8pin connector the cpu uses 12V1+12V2 together.

[Extelleron]

That clarifies things. However I can not follow why JonnyGuru assumes there is no overprotection at all. Looking at this image i'd expect 12V1+12V2+12V3 are overprotected together at 40A max.
Seems you do not have a 18A limit with the PSU. Nice.
It should be a great PSU for the MSI 790FX Platinum which has only one 4pin connector but better PWM than the M3A.
You'll run into mosfet temp issues which such loads on the M3A no doubt.

Another note in the Sapphire/DFI mobo with an 8pin connector the cpu uses 12V1+12V2 together.

Does this board support increasing voltage in BIOS w/ Phenom? Reading through this thread I see that voltage was capped @ 1.25V VDDC in the BIOS for Phenom CPUs.

[justapost]

Does this board support increasing voltage in BIOS w/ Phenom? Reading through this thread I see that voltage was capped @ 1.25V VDDC in the BIOS for Phenom CPUs.

The board uses the cpu's CPU-VID limit. With an BE phenom the CPU-VID is unlocked you can apply up to 1.6V now.
Unfortunately my 9500 does no longer work so I can not revise whether that behaviour has changed with later bios versions than 601.

At the moment the 9850BE is not no their list of supportet cpu's.
http://support.asus.com/cpusupport/c...n-us&model=M3A
An X26400(125W) is supporten. I can not find the latest dual cores 4850e(45W) so that list seems to be out of date.

[thigobr]

The board uses the cpu's CPU-VID limit. With an BE phenom the CPU-VID is unlocked you can apply up to 1.6V now.
Unfortunately my 9500 does no longer work so I can not revise whether that behaviour has changed with later bios versions than 601.

At the moment the 9850BE is not no their list of supportet cpu's.
http://support.asus.com/cpusupport/c...n-us&model=M3A
An X26400(125W) is supporten. I can not find the latest dual cores 4850e(45W) so that list seems to be out of date.

They modified the CPU Support List and now it doesn't have 125W CPUs anymore... I suspect that they won't support Phenom 9850 until a new PCB rev.
That's strange because last week FX62, X2 6000 and 6400 was present in the list.

[justapost]

They modified the CPU Support List and now it doesn't have 125W CPUs anymore... I suspect that they won't support Phenom 9850 until a new PCB rev.
That's strange because last week FX62, X2 6000 and 6400 was present in the list.

Thanks Thigobr,

Yeah I saw at least an X2 6400 (120W) in there as I wrote the above post. Not a bad decision for those buying this board right now. But those whom bought it when those cpu's where in the supported list must feel like beeing tricked.

[informal]

I don't get this new CPU support list thing
The board clearly did support 125W CPUs in the past,and many ppl used the board with heavily OCed 6400+(3.5GHz,even on XS here one user did it),and you justapost OCed successfully your 9850BE on this board!I came across many reports on the web of ppl OCing 6000+ and 6400, Phenoms 9x00 too...Still confused
So what's the deal now??The board still has 5 phase PWM design(4 for cores and 1 for the IMC),why did they remove the 125W duals and quads?Maybe because the board has no active/passive cooling on the mosfets?Maybe they reverted the board design to 4 phase PWM?

[Oc-Ghost]

Or they removed them for marketing reasons, e.g. people should buy their bigger boards

[SocketMan]

The North Bridge heat sink on mine can reach over 80c easily,thats measured on the outside with thermometer. That happens with X4 (B2) and X2,overclocked or not,pcie2 or pcie1 VGA,different (voltage/brand) ram. I've changed the thermal compound as well (to AS5) nothing helps (unless you use after market cooling ).Got a fan sitting on top of the vga card blowing on the NB -my solution.
I am sure they wont mind if we all got another Asus board either

[dball3]

805 is up ftp://ftp.asus.com/pub/ASUS/mb/socketAM2/M3A/0805.zip I flashed and seems to be stable, although i can't really find anything new or different. I disabled the tlb fix through bios and aod still boots up with yellow.

[thigobr]

My M3A arrived yesterday...
My first impression was good until I see that it turn off and turn on after BIOS changes. This new BIOS 0805 solve this problem? This behavior will damage Hard Disks...

[dball3]

For some reason that's the way the bios resets itself, not sure why. Same thing for 805 though. It scared the out of me the first time it did it, i thought it died.

[NEC]

I'm going to buy this for some overclock, any thoughts?

Athlon64 X2 5200
Asus M3A
2x1GB SuperTalent 800MHz

Thanks.

[SocketMan]

I'm going to buy this for some overclock, any thoughts?

Athlon64 X2 5200
Asus M3A
2x1GB SuperTalent 800MHz

Thanks.

Too many........ I would replace the NB heatsink, right away; and be careful with "after market" coolers- some of them (like Enzotech XtreamX) will cover the 1st DIMM what else? Some people had issue with the "onboard" NIC-sudden reboots or BSD's- think the newer drivers solved it. This thread should have "some info'" as well.
************************************************** ************************************************** ************************************************** *****************
Yea that 'shut down" is not long enough, the drives are still spinning when it powers back on, it might be related to the fact that Phenom checks (I think)
during post whether on not it was a "cold reset" and goes into a slightly different "boot routine" for "warm" and "cold' resets.

[CRFX]

I had a bit of a mishap before making me think I killed my M3A.
Have an Antec P182 and the 4pin power wire isn't long enough to run up the back and plug it in, so I got one of these extenders. http://www.newegg.com/Product/Produc...82E16812198005
Worked fine for a while till after I voltmoded my board with 1.5v - 1.55v on my 9500. One morning I wake up to find my system was frozen, so I reboot and get nothing. Look around and faintly smell that something was burnt, so I'm like great..burnt the board. After removing it to replace it with a DFI that I got I find the 4 pin extender cable I used MELTED!! The connector between it and the power supply was black and melted on the 2 12v pins.
So now I cut off the burnt parts and wired the extender to the power supply and it's been working fine. However since I have a DFI M2R now with an 8 pin connecton I'll be using that with a 9850 and use the M3A without the volt mods for now. I haven't encountered any instability but pulling that much from my antec 500w supply with 17A rails probably isn't good for it now that I think about it.