Gigabyte MA790FX DQ6

[Lightman]

Yes, Gigabyte will open up NB voltage points. I am working with AMD right now on the X64 clocking, my systems do the same thing, a good 150MHz lower on Vista64 than I am on Vista32 or XP Pro. I will post up another beta tonight, this one is much better than F3B, not perfect yet, but I think the board outperforms the DFI now.

All good news!
Do you know what is a cause for weaker OC under x64? Is it related to HTT??
Also I'm still having problems with my 2 IBM 40GB 7200RPM ATA drives. They are disappearing on BOOT, but in Windows both are fully functional! If they're not detected on POST I can't boot from them and F12 boot menu freezes instantly... Must be small bug related to SB600 (It was fine with all F1 and F2 BIOSes)

Thanks bingo13 for the F3B bios, I can know boot att 265*11 with NB speed around 2gz+ Working on getting my rig prime stabbel know, Do know know why 3 cores is prime stable at 1,36-1,45v @ 2750 but not the all four? could it have to do with the nb running to high speed? Testing 1750nb\htt know will post results with higer clocked cpu this weekend to!

Very nice results!
To make you fill better, my Phenom core1 is not stable above 2.52GHz stock volts and above 2.64GHz at any volts I set BAD CORE

PS. Here are some 3DMark06 scores which shows CPU score scaling and what F3B BIOS allowed me to achive

[bingo13]

F3C pulled until patched correctly - F3B back up.

http://images.anandtech.com/reviews/...9XDQ6F3BAT.rar

[mAJORD]

mmmmmm So I can't seem to find the option to turn the TLB patch OFF in F3C..

someone tell me they just forgot this time.. or i'm missing it.

[BadNizze]

Very nice Lightman! Need a second card for Cf to catch you know

Finly got my CPU prime stable @ 2750 with 1,392 HTT\NB @ 1750mhz, will try NB at 2000, but prime takes SO LONG time to test...

Cooling is enzothech Ultra X with nexus @ 12V BTW Finly something to meshure temps with to!
http://www.xtremesystems.org/forums/...d.php?t=168545

[mAJORD]

I have issues with prime95 some time after the 10 hour mark on blend. no error, just crashes. (standard 'windows needs to terminate program' error)

[KTE]

Very nice work. Keep it up, I'm always reading.

Just you have better boards/support...

[Lightman]

Another one from the fine people at Gigabyte -

http://images.anandtech.com/reviews/...9XDQ6F3CAT.rar

Thanks again for posting!!

mmmmmm So I can't seem to find the option to turn the TLB patch OFF in F3C..

someone tell me they just forgot this time.. or i'm missing it.

Is this BIOS slower than previous ones??
I will flash it in couple of hours time too

[mAJORD]

it contains the stupid TLB patch, so performance is dismal.

to the point that if the next bios(s) doesn't allow it to be switched off I'm going to pack the cpu in it's box and sell it.. then wait for B3

anyway, hopefully that won't happen.. In the meantime do you have access to F3B by any chance?

[karbonkid]

...if the next bios(s) doesn't allow it to be switched off I'm going to pack the cpu in it's box and sell it.. then wait for B3

From http://techreport.com/discussions.x/13742 -

AMD does mitigate its warnings by saying few customers are actually affected by the bug. However, the company told us yesterday that it is telling motherboard makers to enable the BIOS fix by default and not provide an option to disable it.

[Lightman]

it contains the stupid TLB patch, so performance is dismal.

to the point that if the next bios(s) doesn't allow it to be switched off I'm going to pack the cpu in it's box and sell it.. then wait for B3

anyway, hopefully that won't happen.. In the meantime do you have access to F3B by any chance?

Of course!

I was afraid this will happen, so I back it up

F3B BIOS! Grab it while you can!
DownloadLink: http://rapidshare.com/files/75140844/MA79XDQ6.RAR

Performance hit is worse than terrible On top of that stability is far from F3A or F3B BIOSes. Can't run 2500MHz under XP x64 on newest one at all!
Picture will say it all....


3DMark06 CPU score @2470MHz + Errata = 3075 points (T1 = 1.023FPS; T2 = 1.478FPS)
3DMark06 CPU score @2442MHz no Errata = 3673 points (T1 = 1.214FPS; T2 = 1.777FPS)

HL:Lost Coast 2470MHz + Errata, HD3870 CF @865/2462, 1680x1050, Aniso x8, Highest settings, noAA = 63.63FPS
HL:Lost Coast 2490MHz no Errata, HD3870 CF @865/2462, 1680x1050, Aniso x8, Highest settings, noAA = 142.14FPS

This one is funny (for Intel die hard fans..):


Looks like AMD wrongly evaluated performance hit of this workaround and they claimed 10% instead of 10x

If that is how this CPU will perform then I think AMD should replace all B2 as soon as B3 is available...
Many people can sue them for that....

[muzz]

Yeah, I read a review of the "Patch" yesterday...
Some of the numbers were reasonable, but there were some #'s where it dropped over 50% !!

That's a disgrace.......

[KTE]

The patch messes up everything if you ask me. But it does help quite a bit to smoother running. If benchmarking, it'll decrease perf. Does with MSI boards anyway.

[Hemi]

I can see the lawyers jumping all over this in the US. Reminds me of nvidia"s reference boards they replaced that were defective. I wonder if AMD will be forced to do it or they will recall on there own.

[KTE]

Lightman: Wow. Talk about a bad patch. That is simply scam. You might as well buy an X2 than run a patch that bad.

[Lightman]

Lightman: Wow. Talk about a bad patch. That is simply scam. You might as well buy an X2 than run a patch that bad.

I think K6-III 450MHz would do better in WinRAR than my Phenom @2470MHz

I'm back to my good old bugged Phenom on F3B BIOS It's so much faster!!!! I hope future BIOS rev. will have option to disable that Errata, otherwise I'm selling this CPU....

[bingo13]

I think K6-III 450MHz would do better in WinRAR than my Phenom @2470MHz

I'm back to my good old bugged Phenom on F3B BIOS It's so much faster!!!! I hope future BIOS rev. will have option to disable that Errata, otherwise I'm selling this CPU....

Sorry guys, the F3D had the patch and the F3C was suppose to be the patch without the TLB fix but with the new optimizations. It appears the official F3C I received for release was not patched correctly. I will get a fix from Gigabyte on Monday, provided AMD allows it. Also, we have word the next AOD release will allow the TLB patch to be disabled, however, it's not working in the current beta copy I am testing.

Another item to be aware of is running EPP memory and the system setting memory voltage higher than 2.2V, it appears in testing right now that running memory voltages higher than 2.25V for long periods of time will burn this core up. How do I know this, well let's just say I have killed my last 9600.

[bingo13]

mmmmmm So I can't seem to find the option to turn the TLB patch OFF in F3C..

someone tell me they just forgot this time.. or i'm missing it.

See message above, will get a patched version on Monday if AMD allows it.

[BadNizze]

Thanks for the warning Bingo! i must say thoug over the last few bios memory overclocking has become superb! I´m running Ballisitx 8500 @ 4-4-4-10 2,16v @ 500mhz compleatly stable! Why only 2,3vdimm in bios thoug? My and Ligtmans boars do overvolt so it´s realy 2,43 but still? 2,6 Would be nice for bench Drop me a Pm if you need beta testers for new bioses!

[bingo13]

All good news!
Do you know what is a cause for weaker OC under x64? Is it related to HTT??
Also I'm still having problems with my 2 IBM 40GB 7200RPM ATA drives. They are disappearing on BOOT, but in Windows both are fully functional! If they're not detected on POST I can't boot from them and F12 boot menu freezes instantly... Must be small bug related to SB600 (It was fine with all F1 and F2 BIOSes)

Right now it appears to be memory related. Still working on it but I have narrowed it down to memory. In fact, there are several memory issues we will discuss shortly. I found doing a complete CMOS clear, power up, set defaults, power down, and then power up will bring the Boot Menu and POST detection back after a BIOS update on this board.

[Lightman]

Right now it appears to be memory related. Still working on it but I have narrowed it down to memory. In fact, there are several memory issues we will discuss shortly. I found doing a complete CMOS clear, power up, set defaults, power down, and then power up will bring the Boot Menu and POST detection back after a BIOS update on this board.

I tried that earlier on but still after couple of restarts one or both of my IDE HDDs are disappearing .

I must agree with BadNizzle that memory OC is SuperB!! My Crucials BallistiX DDR2 800MHz C4 can do 1020MHz 4-4-4-12-18 1T with only 2.32V (reading from BIOS) and I've tested them 1182MHz 5-5-5-18-24 2T as well (2.43V though)

Great work on the BIOS Bingo13! I hope you will get that BIOS from AMD

[Brother Esau]

Hey, guys whats you're disk performance in Raid 0 like on the SB600 please?

[Schneider]

Could anyone link me to the version of AMD overdrive they are using? I have ver 2.10 or sumthin and it says no RD700 chipset detected ...CPU-Z says i have the RD780 chipset which i find odd as well.

Board i use is the 790X DS4

[mAJORD]

Lightman, bingo thanks and very good news about the new BIOS coming

[mAJORD]

doin a lot better now with F3B

2.9ghz on water, too many volts required though, and very picky. It's not a fun chip above 2.8

Thanks for the link Lightman

[Lightman]

doin a lot better now with F3B

2.9ghz on water, too many volts required though, and very picky. It's not a fun chip above 2.8

Thanks for the link Lightman

You welcome!

Any test done @2.9GHz??


@Schneider: We all are using AOD 2.10 from www.techpowerup.com ...

I think you will need to wait for new version due soon...


Back to good news about GA-MA790FX!

This is what I achieved today with my Crucial BallistiX 6400C4


My first ever 1.2GHz system memory!! All that with only 2.42VDIMM

Ps. It's stable! I'm typing this post at above setting!