msi 790 fx-gd70 thread

[beta]

APS can switch on inactive power phases depending on load. The problem is, the switching takes time, and during this time the active phases will be inevitably overloaded.

MadDias, something is wrong with your mobo is you cant enable it..

[mav2000]

So would keeping the APS off solve this problem?

[chew*]

cannot handle 4ghz??
i even primed 4.1ghz@1.475v for over an hour and ran vantage at 4.3ghz@1.55v. i don´t know how u cooled ur pwm, but just because it has no fan on the heatsink doesn´t mean that it will run cool without ventilation. and a case fan isn´t really enough to cool the pwm heatsink.
this board can handle a thuban at 4ghz very well...

the board had issues with prime 95 with just a 4 core at 4.2..........I used prime 95 at 4.2 with no fan on PWM to bin boards for ln2.......want to guess how many got RMA'd?

On a side note we had better luck with open box boards as apparently the "bad" mosfets were changed already.

Those not having problems likely just got a "good" board which is what we were binning for........

[MadDias]

MadDias, something is wrong with your mobo is you cant enable it..

i don´t know if it was aps that gave me a bluescreen when booting windows. but since then i never tried enabling something in the "green power" menu again.
i´m pretty sure that´s what killed mav2000´s board. why would u want green power when ur overclocking and running the system at 100%load during prime95 anyways??
my board is rev1 i guess.. the stickers on it say 010 and 10.
the x6 is running with 10w less power usage in idle than my old 965C2 without any "green power" thing enabled.

i don´t wanna guess how many boards went belly up while priming without a fan blowing on the pwm section
but that´s not only a msi problem... i bet every amd board with 4+2pwm screams for mercy when u torture it without giving it fresh air.
that "open box" phenomenon is interesting... so there are really some boards leaving the factory with already borked mosfets.

[drmrlordx]

Ok its dead...so this board cannot handle a thuban at 4 ghz for even benchmarking. Took all of 40 minutes to die. Just saw one small spark and that was the end of it. Im sure the vermont is fried but since i cant see any damage im not going to take out the hs. Going in for rma tomorrow. Just a heads up to anyone else who wants to get a hexa fix.

Ouch man. Something to think about if/when I get the RMA return on mine.

Sorry the board fried. Hopefully the boards they're sending back as returns will be better able to handle Thubans than the boards that are out in the wild?

[vinister]

I have to agree about the PWM temps and the necessity of forced cooling. Using an IR thermometer, I measured temps up to 90 degrees on the PWM chips at 3.8Ghz with an X4. I put a small fan, resting on the top of the graphics card, pointed at the NB/VRM area. The temps now don't break 50, and I was able to push to 4.1Ghz. I plugged the fan into the CPU fan port so I could set it to spin up with load. I have it set to target 40 degrees on the CPU, so as soon as there is any significant load, it kicks up the speed. The lowest speed is set to 40% so its still quiet.

There is no way anyone should be overclocking an X6 without at least some cooling towards the PWM section.

I do use APS because who likes huge power bills? Any little bit helps, I'm not a fan of wasting power if the system doesn't need it. I've always found the APS on this board to work very well and doesn't seem to affect stability at all.

[chew*]

There is no way anyone should be overclocking an X6 without at least some cooling towards the PWM section.

The whole point is to seperate the wolves from the sheep. I would not reccomend it either but good chips aren't extremely common.

For my needs I need to know the board has a rock solid PWM. This is the best way i found to determine it.

If it survives with no cooling in prime95 with a sacrificial chip ( because this board when it nukes has always taken a chip ) then I am willing to put it under ln2 with a good chip with cooling on PWM

[MadDias]

I just received a mail from msi tech support with a link to a beta bios that should solve the "ht sync flood error" when using a hd5970.

BIOS 1.1D4

Now the problem is, I still can´t open multiple cpu-z windows, after the 4th cpu-z window the system got unresponsive and rebooted.
But this time the error message was "Unknown BIOS error. Error code =9200"
I guess they just changed the error message but didn´t solve the problem...

[beta]

So would keeping the APS off solve this problem?

I don't know, maybe the bad mosfets would have died anyway. Yesterday I installed 2x60mm fans on the pwm to prevent this, but tbh i don't want to risk the board and the chip.. maybe I should change to CH4 before I get my thuban (and something bad happens)..

MadDias: thanks for the link!

[=SOC= Admiral]

Im going to have to get a new board because this one is crap for overclocking. Wont let me take my 1055T over 3.6 GHz and even that isnt 100% stable but 3.5GHz at stock voltage is completely stable. Im just going to get the Crosshair IV extreme when it comes out.

[bynext]

i was talking to MSI support and they said * Please don't overclock so much, please adjust your CPU ratio as 18/19 to check again* thats sounds weird and funny coming from MSI..i have some problems overclocking my 1090t and they said that....

[MadDias]

Im going to have to get a new board because this one is crap for overclocking. Wont let me take my 1055T over 3.6 GHz and even that isnt 100% stable but 3.5GHz at stock voltage is completely stable. Im just going to get the Crosshair IV extreme when it comes out.

maybe it´s ur ram.. not the board holding u back.

[MadDias]

received a link from msi today to a modified version of the 1.1D4 bios.
i was hoping that it would solve the stupid reboots when i open multiple cpu-z screens...
but it´s still rebooting... the only thing that changed was, that it´s not showing any error message at post this time.

edit: cpu-z 1.54.1 beta solves the reboot problem!!! would have never thought that cpu-z was the source of the error...

[Oliverda]

received a link from msi today to a modified version of the 1.1D4 bios.
i was hoping that it would solve the stupid reboots when i open multiple cpu-z screens...
but it´s still rebooting... the only thing that changed was, that it´s not showing any error message at post this time.

edit: cpu-z 1.54.1 beta solves the reboot problem!!! would have never thought that cpu-z was the source of the error...

Can you share us the link?

[MadDias]

Can you share us the link?

there u go modified 1.D4

it´s really nothing more than the normal D4 only without the annoying "Unknown BIOS error" message. if u get a reboot now u don´t have to press F1 to continue.

[ridney]

maddias, thanks for the bios link. it seems my 1055T can now clock cpu-nb past 2800+ stable, oc my ram and tightened up cpu vcore, but the problem i'm having right now is that it doesn't run turbo anymore. i can't boot into windows with turbo-on on this 1d4 bios

[ridney]

anyone flashed with the latest 1D4 bios? i am having BSODs with Canon's Digital Photo Professional or even trying to open photos, i get distorted screen and then BSOD saying there's a problem with memory management. RAM is running stock and cpu-nb at 2700, same stable settings i have running previously

[Oliverda]

there u go modified 1.D4

it´s really nothing more than the normal D4 only without the annoying "Unknown BIOS error" message. if u get a reboot now u don´t have to press F1 to continue.

Thanks!



A7577AMS.1C0 ---> A7577AMS.1D3

- Fix max FSB function with 6 cores CPU;
- Fix HT Link Voltage deviation.
- Fix 6 Cores CPU, auto max fsb function sometimes shows oc fail warning message;
- Update AgesaMSI eModule to V3_6.06 (05/19/2010)
a. Modify Unlock CPU Core function to improve unlock success percentage for 960T CPU.
b. Fix system still change PState at idle mode when used Rev E CPU and "AMD Turbo Core Technology" function is disabled.
c. Fix system hang on FFh when used CPU Core Control to disabled CPU core and Unlock CPU Core is enabled.
d. Fix system can't boot to OS when used CPU Core Control function disabled unlocked core then disabled Unlock CPU Core function.
e. Fix the CPU FSB overclock ability not good issue when used DDR3 2GB memory and set "FSB/DRAM Ratio" to 1:2.
f. Patch the "AMD Turbo Core Technology" function's CPUVID too high issue when over voltage from UP6262.
g. Add "Auto" option for "AMD Turbo Core Technology" item and set to default value, the "Auto" option means that Turbo Core Technology will linked to AMD Cool'n'Quiet Technology.
h. Modify the "Adjusted CPU Frequency" and "Adjusted Turbo Core Freq." function code for Rev E CPU.[/list]
- Modify M-Flash boot function will skip CPU FID VID init patch code;
- Update CPU-Z to V0.09.

[ggdh]

Sorry guys for crapping here with such questions but i'm puzzled:

1. All pci-e slots are x16? (when using only one blue slot?)
2. Any reason why when 5870 placed in second blue slot is working at x8??

[vinister]

so what is the latest recommended bios for this board?

Any issues with the latest 1.C on the MSI website?

[zanzabar]

Sorry guys for crapping here with such questions but i'm puzzled:

1. All pci-e slots are x16? (when using only one blue slot?)
2. Any reason why when 5870 placed in second blue slot is working at x8??

all blue slots are physically 16x, if slots 1 and 4 are used then they run 16x if anything is in slots 3 or 5 then all of the 16x slots will run at 8x and slots 3 and 5 are 8x only. then slot 2 is a pci-e 1x going from the SB so it dosnt matter

[ggdh]

Ok, thanks.

Moved the card to third blue slot (PCI-E 4) and it works as x16. Next time i'll have a look in a manual before asking...

[vinister]

so what is the latest recommended bios for this board?

Any issues with the latest 1.C on the MSI website?

So I tried this 1.C bios before installing my SSD RAID-0 array.

Everything works, but now my overclock, which has been stable for months, is no longer stable.

I have been running a phenom 955 under water at 4.0Ghz @ 1.3875V. 200x20. Now that setting is not stable even up to 1.45V!

Why would that happen? I used to be able to run prime95 overnight, I could even run furmark at the same time for hours on end with no crashing. Now it crashes in the first 5 seconds of prime.

Grrrr!

[MadDias]

So I tried this 1.C bios before installing my SSD RAID-0 array.

Everything works, but now my overclock, which has been stable for months, is no longer stable.

I have been running a phenom 955 under water at 4.0Ghz @ 1.3875V. 200x20. Now that setting is not stable even up to 1.45V!

Why would that happen? I used to be able to run prime95 overnight, I could even run furmark at the same time for hours on end with no crashing. Now it crashes in the first 5 seconds of prime.

Grrrr!

why do u need to update to 1.c0 if u had a stable system?

[zanzabar]

why do u need to update to 1.c0 if u had a stable system?

the older raid rom sucks, there needs to an update that boots faster and has better random IO performance