Thanks.
Reason for using 1605 is that I find it more stable with striaght AMD equipment. I get a lot better scores with benches using that BIOS.
Printable View
Thanks.
Reason for using 1605 is that I find it more stable with striaght AMD equipment. I get a lot better scores with benches using that BIOS.
I just noticed some wierdness mixing brands too hah.. The other day, I picked up a XFX 8800GT Alphadog and set it up in the second slot for Physx. The Win7 WDDm driver installed and made it my main gpu:shrug: ! lol.. CpuZ read it as Gpu1 aswell.Quote:
Originally Posted by taz2
http://valid.canardpc.com/show_oc.php?id=1560488
Maybe later I'll swap the gpus and see what happens ..
I don't know if you guys have noticed this or not.
Check the pins on the pcie slots.
Most of my slots only have a portion of the pins there.
Only 2 of my slots have all of it's pins.
Sure there might be specs somewhere listing this or that slot used for 16x and etc...
But it's probably best you check yourself to see what slot is capable of running full speed.
Kind of lame I think, that you can't just pick any slot and run it full speed...
But ohwell.
Better then having a bunch of 1x or 4x slots that you can't fit a full sized card in..., or maybe not ^^ ... I don't know.
Mentioning this because perhaps it may have something to do with the dual card setup probs.
I really hope that 2011 will be a better year for the M4A79.
Or maybe ASUS will let us wait a few more month until they'll say that the M4A79 is too old and won't be supported anymore :D
that's because only 2 of the slots are X16. 2 cards at X16, 3 at X16/X8/X8, 4 at X8/X8/X8/X8
M4A89GTD-PRO-USB3-ASUS-1703
ftp://ftp.asus.com/pub/ASUS/mb/socke...-ASUS-1703.zip
Dated Nov 16. 2010
Quite a long time in validation. Hope this one fixes the USB problems.
M4A89GTD_PRO 1801
ftp://ftp.asus.com/pub/ASUS/mb/socke...-ASUS-1801.zip
Dec 23/2010
Very new bios for non USB 3 board.
the new bios for m4a89gtd-Pro is pretty good
better compatibility for my rams :D
USB still slow via headers on board in new bios for M4A89GTD-PRO-USB3
It was working in pre 16xx bios's I wonder if I should ask bingo13 about it?
Did a quick check on the M4A89GTD-PRO-USB3 1703 BIOS
AHCI ROM: 3.0.7 - (There is at least a v3.0.9 available)
SATA RAID Option ROM: 3.2.1540.6 - (is this to be considered as .06 or .60 version? if .06, there is at least a .35 version available)
I do hope newer versions of the different Option ROMs are included in the upcoming Ch IV F BIOS
(There is at least a v1.19.x JMB36x rom available for example - Asus use v1.07.20)
And yes, i know it takes some time in validation testing for each new option rom that becomes avaliable...
Edit: there is also a NEC/RENESAS USB3 Controller (uPD720200) firmwares Version 3.0.2.7.0.8 available. Is this included in Bios, or would you have to update (flash) it yourself?
I flash the NEC/RENESAS USB3 Controller (uPD720200) firmwares myself in windows.Quote:
Originally Posted by nex_73
Follow instructions, they work fine on our board.
I wish I would have seen this before ! haha.. Adding more ram may fix your problems. A 512mb/1gb gpu needs at least the same amount of system memory to function properly. Running on only 2gb is "starving" them lol.. and its causing your system to rely on the much slower,(will burnout your drive), pagefile.Quote:
Originally Posted by NEOAethyr
Lets hope running the ram soo fast didn't do any damage. gl!
@AmdAtiTom420
My win7 uses about 384megs of ram, my win2k3 about 90megs.
I defently have a ram limitation issue in cod-mw2, the museum level (last, bonus level).
I can tell when I go from one side of the map to the other when enemies are all loaded up, it swaps out for a good 5 secs.
Once in a great while black ops will use the swap file a bit, effecting game play, but it doesn't happen offen.
Does "ok'ish" on 2gigs.
Not the best, but good enough.
I probably did hammer the ram to hard or something.
I'm rma'ing them monday morning.
Hopefully that is my prob, bad ram (still unstable at stock speeds, even for mild tests).
The board concerns me alot though.
It's the lights.
The ones for the power button stuffs.
I shut my computer down, and those don't stay on like they they used to at all.
I hope it wasn't this psu's 5vsb that shorted it out and messed up my board.
That would really suck.
I've never took the time to measure the 5vsb on this psu... :rolleyes:
Hopefully it's just acting up in odd ways because of my ram ^^.
I guess I'll find out in a few weeks :).
cool and quite doesn't work at all on 1101 for M4A89TD usb 3.0/pro.
can't acces it or select it nor can I change C1E now either :/
I noticed this too :(
Are you setting CPU & NB Voltage Mode to Offset?. C1E and C n' Q are locked out if set to Manual.Quote:
Originally Posted by demonkevy666
Hey guys!
I've just bought the ASUS M4A89GTD-PRO/USB3 :D
So what is the best BIOS of this mainboard for overclocking with 1055T at this time?
Thanks a lot :D
//Sorry my Eng, I'm Vietnamese.
At last, M4A89GTD BIOS 1703 detects my two Geforce cards! (460 & 240)
New Year's Eve gift from ASUS. Thanks!
But...why.......Quote:
Originally Posted by Praz
Now I wish I hadnt flashed from 0901...
Any news regarding the latest CH IV Formula BIOS?
+1, I've worn out the F5 for over a month now... :yepp:Quote:
Originally Posted by Oliverda
ok it working now, earlier I tried both but nothing happened -.-Quote:
Originally Posted by Praz
it's working right now which odd lol
Quote:
Originally Posted by Ccaution
Most of us have been waiting for a stable bios since release ! :mad:
You're lucky, Radeon+GF still undetectable, system see only AMD card...Quote:
Originally Posted by oxydius
How much ram are you running ? How much ram do the cards have?Quote:
Originally Posted by -Px-
My only prob running ATI/Nvidia is, Windows will make the 8800gt the main display even though its installed in the 2 pcie slot lol. Then yesterday I tried messing with it again. This time, Windows picked the 5750 as the main display:up: but forgot the desktop icons over on the 2nd monitor/8800GT. :down:
CpuZ read it the same as well. Physx works great, btw. :D
5750 + Physx
http://img.techpowerup.org/101228/Ati physx 1.jpg
XFX 8800GT 512mb solo run
http://img.techpowerup.org/101228/8800 GT solo.jpg
Edit
CpuZ updated :)
Filled specs in signature, but as I see in your sig, you have M4A79XTD EVO mobo, which is completely different case...Quote:
Originally Posted by AmdAtiTom420
This problem seems to be consistant with many Asus mobos. And besides.. our chipsets are almost identical lol... They added USB3/Sata3 and rebranded it. :DQuote:
Originally Posted by -Px-
Still no new BIOS for CH IV Formula...:(
Can anyone double check the crosshair 4 for cpu ecc...
On mine it said not enabled when I checked with aida64 and everest.
I'm starting to think it might not been the program.
But I didn't check by hand with wpcredit or anything.
I'm thinking it don't work and that might be a good thing to have working.
I don't care so much about mem ecc, just cpu ecc, I want to use this...
( Post a ss of what you want to check, and I'll post back the same on my rig ;) )
Oliverda: this days are days of Sandy Bi*ch :), so... :-/
can anyone else confirm the M4A89GTD Pro/USB3 now correctly recognizing two nvidia cards at the same time? i will do my own test with a GF104 and G92 tonight, thanks!
I would like to request an improvement to the M3A78-EM bios: CPU/NB frequency multiplier.
Actually, Asus put it in, then took it out early on. Why ??
I know NB multiplier tweaking is supported by the board, because I can adjust it with BarEdit within Windows. But it would be much more stable if the NB multi could be saved in the bios, rather than loaded on Windows startup with Baredit.
http://www.xtremesystems.org/forums/...1&d=1294299575
I see there was a new bios for the Asus M4A89GTD PRO/USB3
And the notes are 'improved system stability'
Does anybody already know if the unstable HT bus ( mine is on 300 but keeps changing to 301 and back etc ) Same with NB, all my speeds are a bit random moving between 10~20MHz. Does anybody know if that's fixed with the new bios?
Else i don't see a reason to upgrade to the new one.
confirmed the 1703 bios for the M4A89GTD Pro/USB3 now recognizes two NVIDIA cards at once (GF104 and G92):
http://i.imgur.com/YDx2z.jpg
hey bingo.
whats up?
Doubt he will be posting in the next couple of days with CES going on.Quote:
Originally Posted by pershoot
There is a BIOS ver. 1304 available for the Crosshair IV Formula on the Taiwan FTP site.
Thanks dloading now
Some new things were added
http://img288.imagevenue.com/loc92/t...5_122_92lo.jpg
The SATA II Host controllers and the Raid ROM wasnt updated
but what do these new settings do :shrug:
Is there a changelog to be found somewhere on the net ?
New BIOS Updates on First Page - More Coming this Week.
SATA RAID Option ROM: 3.2.1540.6 - is this to be considered as .06 or .60 version?Quote:
Originally Posted by bingo13
PCI Express to SATAII HOST Controller ROM v1.07.16
AMD AHCI ROM 3.0.5
I'm somewhat dissapointed your guys haven't take the time to update the different option roms in this one :mad:
Can you kick them in the b*tt until next beta release please.
Now of to flash & test this new one :up:
IV Formula - 1304 update done, first thing I noticed is that is doesn't load my OC profiles. The profiles are there to be seen but once you try to load one, you get an error message :shrug:
i have never been able to load profiles across differant bios's with asus boards, so dont hang up on that, any other news?
I did load profiles on other asus boards (i.e. m4a78 pro) across bios updates - I'm testing 50% LLC at the moment, MM'd too - smooth voltage transactions there, seems nice :)
where is this bios bingo? or its previous version 3702?
http://img225.imageshack.us/img225/7686/image1cs.jpg
its been since mid june last year since this board has had an update :yepp:
Bingo13 ,what about older boards? M3A79-T deluxe.Any news about some new bios ?we stiil need a Turbo OFF function for them.
Can we get some explanation regarding the new voltage related functions?
http://prohardver.hu/dl/upc/2011-01/13063_asus.png
Anyway with the default settings I get exactly the same overvoltage issue which is present in the more than 6 months old 0905.:shrug:
I will be in the office in a couple of days and will post up the changes on the voltage settings and what they are intended to accomplish.
Any chance the AHCI verison of any of the bios's will move to the latest version?
The asrock boards get the new versions with there bios's
Any chance the AHCI verison of any of the bios's will move to the latest version?
The asrock boards get the new versions with there bios's
I will be back in the office tomorrow and will have a listing ready of the various voltage settings.
Can you also address the throttling issue I have been seeing after flashing from 055 to 1304. Seems like when the processor hits 52* it drops to 1000 MhzQuote:
Originally Posted by bingo13
CPU/NB voltage OPC? What is it?
Haha...same problem I've got.Quote:
Originally Posted by Dirk Diggler
Where is that M4A79 Deluxe bios? It's been in testing a long while, and it haven't been any beta release in a long long time...
I have still yet to experience this on the M4A89GTD Pro or the Crosshair with any bios.Quote:
Originally Posted by Dirk Diggler
Is there any afudos version which will allow me to flash back to 0905? I tried several version, but it says smth like "ERROR: Bios has no flash information available".
I'm using
where bios.rom is the filenameCode:afudos /ibios.rom /pbnc /n
PS: Tried 207, 211, 226, 229
AFAIK you just can't flash back to an older bios.Quote:
Originally Posted by I.nfraR.ed
Nice to see new options in the bios for the ch4.
Not quite what I was expecting but good stuff.
It really would be nice to see multipliers, and vid's for each power state mode.
The missing mem timings.
And working ecc for the cpu.
Cpu-nb config stuff, like unit id clumping and clock modes.
Pcie config stuffs like speed, pcie link state power management config.
And the works, ie all latencie reg's for every chip on the board.
Including 2nd'ary latencies and disgard timers, stuff like that.
And....
One more thing I can think of to top it off.
tAC and tCK timings for all 4 slots, only need 1 set of timings, rather not run my mem at diff timings when it comes to those...
I can mod the spd's, but I'de rather not these days.
Modding it with the bios, temp like for each boot, would be freaking awesome.
Would allow for a slightly lower latency and slightly higher max clocks.
Oh shoot one more duh...
The master pic controller, the 8253/4.
I would like to beable to change this timer to something like 1.5x the normal speed.
Example, "POST_67S" in award bios'es.
Sorry for spamming:
2nd section is the code I usually use, 1st was old and for some other board at the time I last messed with it.Code:;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
;; POST Table 67 (3rd LED - Boot Logo (Possibly Well Before 3rd LED)) [POST_24S]
;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;
if Board_ID = "6A61BD49" & Cleanup = 1
POST_67S: ;Test 8259 Functionality
cmp byte [bp+0x0246], 0x01
jz near Test_8259_Fun_ret
POST_67S_Continue: ;NEW
cli ;Disable Interrupts
mov ax, 0x0000 ;G_RAM, seg_0
mov ds, ax
;Backup RTC Vector, Move INT 08h vector to a temporary routine
mov dx, [0x0020] ;int08, Save Vector
mov bx, [0x0022] ;int08+0x0002
mov word [0x0020], Test_8259_Fun_Wait ;int08, Timer ISR Entry
mov [0x0022], cs ;Code Segment
;Enable just the RTC Interrupt
mov al, 0xFF ;Mask all interrupts on controller 2
out 0xA1, al ;B8259+1
dec al ;Mask all interrupts except for IRQ 0x08 (tCLK)
out 0x21, al ;A8259+1
;Set Up tCLK/tCTC0
mov al, 0x30 ;Count Generation I Think
out 0x43, al ;8254 CTL
mov ax, 0x0008
out 0x40, al ;tCLK LSB
dw 0xC48A ;tCLK MSB, mov al, ah
out 0x40, al
;Wait For Interrupt
dw 0xC033 ;xor ax, ax
mov cx, 0x7000
sti
out 0xEB, al ;Begining of Loop
out 0xEB, al
out 0xEB, al
loop $-6 ;Loop to 1st Delay
cli
;Disable Interrupts
push ax ;Save Interrupt Counts
mov al, 0xFF ;Mask All Interrupts
out 0x21, al ;A8259+1
mov al, 0x34 ;Control (tCLK), Rate Generation
out 0x43, al
; mov al, 0x74 ;Get the CMOS Value
; out 0x74, al
; in al, 0x75
; and al, 0x0F
;
; cmp al, 0x00
; je Set_tCLK_1X
; cmp al, 0x01
; je Set_tCLK_1125X
; cmp al, 0x02
; je Set_tCLK_125X
; cmp al, 0x03
; je Set_tCLK_1375X
; cmp al, 0x04
; je Set_tCLK_15X
; cmp al, 0x05
; je Set_tCLK_1625X
; cmp al, 0x06
; je Set_tCLK_175X
; cmp al, 0x07
; je Set_tCLK_1875X
; cmp al, 0x08
; je Set_tCLK_2X
; cmp al, 0x09
; je Set_tCLK_25X
; cmp al, 0x0A
; je Set_tCLK_3X
; cmp al, 0x0B
; je Set_tCLK_35X
;
; Set_tCLK_4X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x40 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_35X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x50 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
Set_tCLK_3X:
mov al, 0x00 ;LSB
out 0x40, al
mov al, 0x60 ;MSB
out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_25X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x70 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_2X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x80 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1875X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x90 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_175X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xA0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1625X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xB0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_15X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xC0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1375X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xD0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_125X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xE0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1125X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xF0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1X: ;Reset tCLK, Result is 0x36 tCLK-CTL, and 0x0000 tCLK (14 Bytes)
; mov al, 0x36 ;Square Wave
; out 0x43, al ;8254 CTL
; dw 0xC032 ;xor al, al
; out 0x40, al ;tCLK LSB
; out 0x40, al ;tCLK MSB
;
; Set_tCLK_End:
;Restore RTC Vector
mov [0x20], dx ;int08, Restore Old Vector (int 8 vector)
mov [0x22], bx ;int08+0x0002
;Check For EOI
pop ax ;Restore Interrupt Count
dw 0xC00B ;Any Interrupts?, or ax, ax
jnz Test_8259_Fun_ret_2
stc
jmp Test_8259_Fun_exit
Test_8259_Fun_ret: ;New Code, only active on a 100% cleared CMOS (Errata #1)
mov si, Msg_Errata_1
call DISP_ERR_MSG
jmp POST_67S_Continue
Test_8259_Fun_ret_2: ;Original "Test_8259_Fun_ret"
clc
Test_8259_Fun_exit:
cli
mov al, 0xB8 ;Init Controller 1
out 0x21, al ;a8259+0x0001
;Mask Channel 4 Off
mov al, 0xBF ;Mask IRQ 0x0009 (Keyboard ?)
out 0xA1, al ;b8259+0x0001
sti
ret
else if Cleanup = 1
POST_67S: ;Test 8259 Functionality
cmp byte [bp+0x0246], 0x01
jz near Test_8259_Fun_ret
POST_67S_Continue: ;NEW
cli ;Disable Interrupts
mov ax, 0x0000 ;G_RAM, seg_0
mov ds, ax
;Backup RTC Vector, Move INT 08h vector to a temporary routine
mov dx, [0x0020] ;int08, Save Vector
mov bx, [0x0022] ;int08+0x0002
mov word [0x0020], Test_8259_Fun_Wait ;int08, Timer ISR Entry
mov [0x0022], cs ;Code Segment
;Enable just the RTC Interrupt
mov al, 0xFF ;Mask all interrupts on controller 2
out 0xA1, al ;B8259+1
dec al ;Mask all interrupts except for IRQ 0x08 (tCLK)
out 0x21, al ;A8259+1
;Set Up tCLK/tCTC0
mov al, 0x30 ;Count Generation I Think
out 0x43, al ;8254 CTL
mov ax, 0x0008
out 0x40, al ;tCLK LSB
dw 0xC48A ;tCLK MSB, mov al, ah
out 0x40, al
;Wait For Interrupt
dw 0xC033 ;xor ax, ax
mov cx, 0x7000
sti
out 0xEB, al ;Begining of Loop
out 0xEB, al
out 0xEB, al
loop $-6 ;Loop to 1st Delay
cli
;Disable Interrupts
push ax ;Save Interrupt Counts
mov al, 0xFF ;Mask All Interrupts
out 0x21, al ;A8259+1
mov al, 0x34 ;Control (tCLK), Rate Generation
out 0x43, al
; mov al, 0x71 ;Get the CMOS Value
; out 0x74, al
; in al, 0x75
; and al, 0x0F
;
; cmp al, 0x00
; je Set_tCLK_1X
; cmp al, 0x01
; je Set_tCLK_1125X
; cmp al, 0x02
; je Set_tCLK_125X
; cmp al, 0x03
; je Set_tCLK_1375X
; cmp al, 0x04
; je Set_tCLK_15X
; cmp al, 0x05
; je Set_tCLK_1625X
; cmp al, 0x06
; je Set_tCLK_175X
; cmp al, 0x07
; je Set_tCLK_1875X
; cmp al, 0x08
; je Set_tCLK_2X
; cmp al, 0x09
; je Set_tCLK_25X
; cmp al, 0x0A
; je Set_tCLK_3X
; cmp al, 0x0B
; je Set_tCLK_35X
;
; Set_tCLK_4X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x40 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_35X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x50 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
;; Set_tCLK_3X:
;; mov al, 0x00 ;LSB
;; out 0x40, al
;; mov al, 0x60 ;MSB
;; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_25X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x70 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_2X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x80 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1875X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0x90 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_175X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xA0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1625X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xB0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_15X:
mov al, 0x00 ;LSB
out 0x40, al
mov al, 0xC0 ;MSB
out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1375X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xD0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_125X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xE0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1125X:
; mov al, 0x00 ;LSB
; out 0x40, al
; mov al, 0xF0 ;MSB
; out 0x40, al
; jmp Set_tCLK_End
;
; Set_tCLK_1X: ;Reset tCLK, Result is 0x36 tCLK-CTL, and 0x0000 tCLK (14 Bytes)
; mov al, 0x36 ;Square Wave
; out 0x43, al ;8254 CTL
; dw 0xC032 ;xor al, al
; out 0x40, al ;tCLK LSB
; out 0x40, al ;tCLK MSB
;
; Set_tCLK_End:
;Restore RTC Vector
mov [0x20], dx ;int08, Restore Old Vector (int 8 vector)
mov [0x22], bx ;int08+0x0002
;Check For EOI
pop ax ;Restore Interrupt Count
dw 0xC00B ;Any Interrupts?, or ax, ax
jnz Test_8259_Fun_ret_2
stc
jmp Test_8259_Fun_exit
Test_8259_Fun_ret: ;New Code, only active on a 100% cleared CMOS (Errata #1)
mov si, Msg_Errata_1
call DISP_ERR_MSG
jmp POST_67S_Continue
Test_8259_Fun_ret_2: ;Original "Test_8259_Fun_ret"
clc
Test_8259_Fun_exit:
cli
mov al, 0xB8 ;Init Controller 1
out 0x21, al ;a8259+0x0001
;Mask Channel 4 Off
mov al, 0xBF ;Mask IRQ 0x0009 (Keyboard ?)
out 0xA1, al ;b8259+0x0001
sti
ret
end if
The function "Set_tCLK_15X" (1.5x), is about what I want it at...
This would speed up booting ALOT...
And making things a tiny tiny bit more snappier overall.
Should be adjustable though, because it causes some peep's probs with the drives.
The prob, drives use this timer to setup there internal timings of there own.
If it's to fast, the drive can't work.
So it should be adjustable in steps, default 1x or 0x0000/0000h.
That's pretty much what I want...
But I doubt anyone cares lol... ^^
I guess it doesn't hurt to put my req out there once though...
I'll try the new bios "if" my setup is stable after I get my mem back from rma...
Should be anyday :\.
Edit:
Maybe I shouldn't be posting code like that..., ohwell.
Actually it's OCP. I assumed it meant "OverClock Protection" so I turned it off for the CPU. (see Below)Quote:
Originally Posted by FlanK3r
I flashed 1304 and I believe I experienced throttling for the first time. All temps were in acceptable ranges according to HWMonitor....Quote:
Originally Posted by Dirk Diggler
I was pushing for a CB11.5 run with a score of 7.90. I raised the HTRef 1 Mhz and all of a sudden my score dropped to ~ 5.56. :(
CPU-Z and TurboV still showed my core clock unchanged (~4.5Ghz), but something definetly happened that I haven't experienced before.
Nothing I did would get it back in-line short of rebooting. And this was done with OCP disabled...
Bingo13: Any info on the new Bios settings?
I appreciate the steppings for LLC, but they don't seem to function the way one would expect them too... :shrug:
For instance, setting LLC to 50% actually results in Vdroop of ~.022v when the core is loaded.
I was expecting it to add to core voltage under load (less than 100% LLC of course), but not to lower Vcore!
FWIW, I've gotten the best results with these (Extreme Engine) settings using 1304:
http://i30.photobucket.com/albums/c3...t/DSCN0548.jpg
thx Dave for the info!
Over Current protection ? darn can't test the bios, board is somewhere else at the moment For your LLC findings Dave, seems Asus added the same LLC values as with Sandy Bridge. 50% Calib means you still get Vdroop, 100% should try to keep it stable as the bios setting...Quote:
Originally Posted by Daveburt714
@bingo13
you said some month ago, that there won't be such a big improvement regarding voltage stability for the 790FX boards (unlike the changes for 8XX boards).
So why does it take more than a half year to release a new bios for the M4A79 Deluxe? :shrug:
Because the 8 series boards and intel boards are more important right now:confused:Quote:
Originally Posted by diddi35
beep your not helping my hopes on getting a new bios lolQuote:
Originally Posted by BeepBeep2
Then you shouldn't promise anything if you cannot keep it :welcome:Quote:
Originally Posted by BeepBeep2
Sure, and already was over a half year ago.Quote:
Because the 8 series boards and intel boards are more important right now
At least bingo13 should have said "sorry, but we don't support mainboards that are older than 1-1,5 years" instead of saying again and again that there will be something "very soon".
ASUS = pay more - get less ? :confused:
July 2007, I made a decision to go with AM2 vs LGA775 and conroe, I had $1,700 to spend on a rig and conroe was just slapping Windsor and Brisbane all over the place. The only thing that helped me with my decision was that AM2+ stars cores and potentially AM3 Deneb cores would work in my board. I sort of made a mistake because one year after I was promised support, Foxconn dropped BIOS support for this board mid-2008 without any warning at all.Quote:
Originally Posted by Rolle2k
Actually they just seem to be slow. Even the M2N32 SLI (5 years old) board supports X4's up to the 970.Quote:
Originally Posted by diddi35
Ok lets wait another 4 and a half year for a new bios :DQuote:
Originally Posted by BeepBeep2
(And while waiting I'll get another non ASUS board for sure ;) )
And before ASUS starts to support their cheapest and oldest OEM boards (even the M2NPV-VM got a beta bios for AM3 CPUs) they should care a little more about the "premium" products.
diddi35 belive me i doesn't worth to wait for the BIOS on your board ,maybe eventually they will release a bios that will support the AM3 CPU but like on my board it will be very very buggy.( 1090T ones )
Turbo Core technology not working like it should / no option to disable it like on other AM3 board series, so if u want your board to support a CPU that works like haveing Cool'n'Quiet and C1e and Turbo core all in same time for all cores not only for the 3 of them and u cant disable or enable CnQ or C1e cose it seems they are locked.
Better wait for a AM3+ board and with that u will be abble to use a AM3 CPU and a AM3+ CPU.
Dont know how others are but Asus it's very slow in posting updates for it's motherboards and if they are older motherboards thats even worse.
I have a M3A79-T Deluxe board and it works great with older CPU like 940 / 945 / 955 / 965 but it can't do the same with new CPU , with those are very very bugy.So now i keep the 1090T for safe keeping untill i can get a AM3+ board and i'm sorry to say this but it wont be Asus board.
I don't see why you guys complain so much about asus in general.
Alot of there older boards at least support the newer cpu's in one way or another, better then a no post.
This thread exist, at least they attempt to listen to the users and provide us with the betas.
And besides, if you got such and old setup, and the cpu don't work, then buy a new board and some mem to go with it for your new cpu.
Good boards are not cheap these days though.
But that's just the way things are.
I'm poor as hell but I got myself a new setup to replace the old dfi I had, because it didn't support the 1090t that I needed for gaming and emulation.
If it's an am2...
There is probably not a single award bios baised board out there that even remotely supports it so count yourself lucky it even boots with an older ami bios ^^.
Quote:
Originally Posted by NEOAethyr
You have to know , Asus is not supporting this tread . Bingo13 does it in his free time , and only because he wants to .
@NEOAethyr
So you mean that the M4A79 Deluxe for example (the AM2+ version of the M4A79T Deluxe) is already old? :rolleyes:
And that board wasn't a cheap board. Why should I bother to buy an ASUS board as my next board if I know that the bios support for it might end one and a half year after its release?
I got a M4N82 Deluxe to the side I got from a buddy last summer for building him a new am3 system.
He gave it to me because his computer fell over, from a table to the floor and his drive bay wasn't attached to anything, it ripped 2 of the sata ports out of the board, leaving the pins.
He still had the plastic cover for it.
I did some reasearch the next day, seemed like a decent board.
I called him up, asked if I could come over and get that cover, thinking I'll carefully slide it back on, he told me he tossed it :\.
Not sure if I should pull the pins or rma the board.
Board is working though, I know it posts fine and the other sata ports work.
I'm looking at asus'es site right now, newest bios: 2010/12/16 ^^.
Actually it looks like this board is updated alot....
Surprised me when I noticed it supports the 6 core cpu's (might come in handy if my mem comes back and my setup is still not stable...).
I think if you can get your point across, they will fix the probs you have with the bios.
It appears to me that they are still supporting some of these older boards.
And when I say old boards, I mean am2.
Try looking for good ddr2 mem on newegg, 99% of it is gone.
I sure as heck wouldn't want to go looking for some decent ddr2 right now...
That's why I say, these am2's are old.
It's hard to get good mem for them now.
Right now ddr3 is in it's prime, lots of good stuff.
The 1 gig sticks are getting to hard to find, but 2gig sticks are half the price that they were last summer.
And they aren't your 1333mhz 8 timing bs either, you can can get some decent stuff for reasonable price.
I don't see bingo around this thread much anymore though.
I kinda wish that the engineers would appear in the thread as well to help out.
Now if they can continuesly update a nvidia board, which is is a very very old chipset regardless of what it's labled.
Then there is no reason why a 790fx shouldn't be supported...
But after a year and a half, these board's, the supports goes downhill.
That's the way it is with all brands.
You might get a small update that supports a new cpu, but the usability of that update well, ^^...
One more year and the ch4's will be in the same boat.
Might even be quicker if the am3+ boards come out and the cpu's.
Then we can all say bye bye to good bios updates for these...
I doubt they will support the am3+ cpu's if there ends up being a 192 or 256bit mem controller.
I knew that though, I just wanted a good board to go with a decent 6 core cpu.
Bah, now I'm rambling...
Sorry about the delay, about half of us came down with the flu after CES, not a good time last week. ;) I spoke with the lead engineer on the C4F this morning and he will have the BIOS change log and technical descriptions for the new voltage options tomorrow. We also expect additional BIOS releases this week on the AMD side.
Hey bingo, thanks for the update. Curious to know: Does the new C4F beta allow booting from PCIe?
bingo13
Some news on M4A89GTD PRO/USB3 on ATI+NVIDIA cards working together, please :)
good news Bingo.
The C4F already has this capability. There are issues depending on what is being used but there are also workarounds.Quote:
Originally Posted by hokiealumnus
Quick Overview of Critical 1304 BIOS Settings -
PWM Phase Control:
- This option allows user to choose whether or not the CPU VRM power phases should change dynamically with respect to the change of CPU load. This option along with CPU and CPU/NB LLC will allow the user to fine tune the PWM and CPU voltages for improved overclocking or limiting voltage overshoots during overclocking.
CPU Voltage OCP:
- This option allows the user to disable the “Over Current Protection” of the CPU. Enabling this option may allow the CPU to consume more current than the safety range defined by AMD, but might assist in high overclocks, especially under extreme cooling. Please treat this option with extreme care to prevent permanent damage to the CPU.
CPU PWM Frequency:
- This option allows user to override the switching frequency of the CPU PWM controller. Raising this option may assist in ensuring power delivery is more stable under overclocking conditions but with the cost of higher operating temperatures of the CPU PWM controller.
CPU Load-Line Calibration:
- This option allows user to decide whether or not the motherboard should lower the CPU Vcore as the current consumption of the CPU voltage raises to meet the CPU load-line spec defined by AMD. Users can now choose other levels of Load-Line Calibration in the 1304 BIOS to lower the amount of voltage the Vcore is raised up to or over normal load line voltages as CPU loading increases. Most air/water overclocks should use 50~75% levels for maximum clocks along with tuning of the VRM options. Due to the design of the AMD Phenom series, realize that certain voltage overshoots will be present in Core 0.
CPU/NB Load-Line Calibration:
- This option allows user to decide whether or not the motherboard should lower the CPU/NB voltage as the current consumption of the CPU/NB voltage raises to meet the CPU load-line spec defined by AMD. Users can now choose other levels of Load-Line Calibration in the 1304 BIOS to lower the amount of voltage that CPU/NB is raised up to or over normal load line voltages as the CPU/NB loading increases.
CPU/NB Voltage OCP:
- This option allows user to disable the “Over Current Protection” of the CPU/NB (i.e. the memory controller). Enabling this option may allow the CPU/NB to consume more current than the safety range defined by AMD. Please treat this option with extreme care to prevent permanent damage to the CPU.
Working on a tuning guide with a retail 1100T now.
Thanks for the detailed notes bingo!
Thank you for that; indeed it did boot fine. :up:Quote:
Originally Posted by Praz
I'm a bit confused about what you said bingo.
OCP, when this is enabled does it mean that "Over Current Protection" is enabled?
Because they way you wrote it, it makes me think it's the oppiset.
Other ?.
Load-Line Calibration.
So if we disable this, it means we get more stable voltage if the amp draw increases?
I thaught enabling it did that....
I'm not trying to confuse everyone.
I just want some clear answers lol.
I got one more.
Note my board is offline for the moment, waiting for mem.
If I jack the pwm freq all the way up, as long as I'm actively cooling the pwm it will be ok right?
(edit: does pwm freq effect more then cpu vcore pwm speed?, like cpu-nb pwm...)
;;;;;;;
This ? goes to anyone with the know how...
Has anyone messed up one of these x6's by up'ing there cpu-nb voltages to high?
thx bingo for that.
i can only assume then that OCP enable means just that, leave the protection enabled (default settings)?
can i assume also that motherboard/system temp is the PWM temp? i haven't seen this raise much by setting 500KHZ (maybe 1-2c during load). i do have a low rpm fan blowing from top, but it isnt doing too much due to the way my case is layed out, but it is getting some airflow (i have a fan blowing straight on the nb area).
so far testing on the bios seems good. LLC set to full calibration. i was able to lower voltages a small bit. seems not to waiver as much as the previous. ref. clock sways as it did with the previous.
clock: 4g, NB: 3g, HT: 2g.
thx bingo!
OCP DISABLED disables Overcurrent protection.Quote:
Originally Posted by NEOAethyr
Most common place that you would run into OCP is on ln2 running say wprime 1024m.
Symptoms of OCP is a complete powerdown where board will not fire back up without the PSU being cycled on and off.
If you think OCP is kicking in on air/water/phase/DI you are surely mistaken, it takes over 6 gig all cores loaded to trigger OCP in my experience.
I highly suggest if you do not experience these symptoms that you do not disable OCP.......OCP protects your hardware.
Commonly people hard mod boards and or video cards for this, in fact I originally requested a mod only after speaking with some reviewers who tested on ln2 however ASUS chose to do it via bios.
In most cases overclockers will not notice OCP on air/water/phase/DI.
Yes it is very easy to mess up an X6 with excessive cpu_nb volts.
I reccomend for safety no more than 1.35 on air......
LLC enabled is out of SPEC for AMD cpu's, intel for that matter to. Use it with caution, its always better to have a little droop than none at all.
I have experimented with the PWM frequency on cold and normal temps, I really saw no major gains and reccomend auto for most users.
My personal reccomendations for 24/7 users is 50% core only LLC and leave the rest auto only on thuban, its not needed for deneb.
My personal reccomendation for ln2 users is 100% llc cpu, 50% llc CPU_NB and OCP disabled for NB and CPU, if you must for whatever reason feel the need for PWM run 500K while on ln2...............
I'm guessing the DMM had shown volt spikes? :confused: I'd rather be able to run a lower voltage at idle and a slightly higher voltage when my CPU needs it but if LLC is dangerous in any way I'll disable it...(which will really hurt my OC)Quote:
Originally Posted by chew*
LLC just bumps the voltage up (effect, I don't really know how does it calculate how much), so if you are going balls to the wall to begin with...
Thanks Chew for the quick guide there..will try these soon.
Its out of spec but in this particular boards case it's kind of a necessity with thuban, the voltage regulation just dips to much.Quote:
Originally Posted by BeepBeep2
50% LLC kinda puts the board in a more normal droop state, I found 100% to overshoot more than my liking at least in air/water load situations.
LLC doesn't just bump the voltage in general, it tends to bump it up on demand aka Load, thus the name Load Line Calibration.
Thanks for your insight chew*. :up:Quote:
Originally Posted by chew*
I don't claim to have alot of knowledge conserning electrical properties, but with my brief testing with 1304 it does seem to run more stable with OCP disabled on the CPU...
I'm not pushing crazy voltages (~1.45 core/1.265 IMC (OCP enabled on IMC)), and I only take the clocks to where they stop scaling.
Do you really think I'm risking damage to my chip just by disabling OCP without pushing crazy v's? :shrug:
Like I said my knowledge of electronics is limited, but as long as I'm being sane with my OC/voltages, current shouldn't be a problem, should it?
I'm not questioning your skills, just asking for clarification.... :p:
chew, think, do u mean 1.35V CPU/NB V for 24/7, right? Thx.
bingo13
Don't know, if it already found/reported, but there is bug in 1606/1703 bioses for M4A89GTD PRO/USB3 motherboard: if 8 GB (4x2 GB) RAM installed and chipset SATA controller disabled, Windows 2008 R2 setup/boot stops with BSOD with such text
With only one memory stick, or chipset controller enabled, bug doesn't appearQuote:
The bios in this system is not fully ACPI compliant. Please contact your system vendor for an updated bios.
Technical information:
*** STOP: 0x000000A5 (0x00000000001000, 0x0000000000000000, 0x00000000FFFFFFFF, 0x0000000000000200).
PWM on auto disabled CPU OCP
System rebooted almost immediately on 50% Load Line Calibration when running Prime95.:shrug:
So ? you need to up the Vcore. Don't see the problem...
There is no problem, with LLC on full the system completed 27.5 hrs prime95 testing max temp 47c.
vcore is on 1.425v
CPU/NB on 1.17
I would rather keep the vcore low and let the LLC handle the voltage.
Quote:
Originally Posted by Daveburt714
OCP disabled on air/water making things more stable honestly in my opinion is more placebo effect.
This is the way OCP works.......lets take a hypothetical number.
Lets say the current OCP kicks in at 150 amps draw.......that would be equivalent to about a 400 watt draw just from the socket which i doubt anyone is hitting even on DI.......
unless you exceed that draw and the board shuts off completely.........not a crash that allows you to reboot. A hard power down with no response from power or reset buttons........only then is OCP effecting you. At that point you would have to cycle PSU on and off just to get board to post again.
That is how OCP works, it doesn't make an overclock air/water OC more stable, what it allows is you to possibly damage CPU or board by drawing what would be deemed unsafe amount of current thus the reason there is an OCP limit to begin with.
The current OCP limit without it disabled kicks in about 6gig in wprime 1024 with a 600 watt draw from the socket which if my killawatt still worked I could give you an amperage draw number however i broke it.
As for LLC 100% versus 50%.......well obviously you will need to use a higher vcore with 50% llc but the droop is at least still in the safe zone where no droop can cause permanent damage.
What do you mean no droop can cause damage?
You disable llc for the cpu, like in the older bios even, and voltage level's are lower.
But there is less flutuation.
You mean to say this can damage the cpu in some way? (even at lower levels?)
I ran my board like that for 2 days.... :|
And that was RIGHT around the time my setup started to act up...
Pretty darn close, 48hrs within the time my setup was acting up "I think".
If not maybe 2 weeks absolute tops, but still close to the time my system became completely unstable.
I can understand the 50%.
With it you could keep a halfway low vcore level, and the droop would be less then 100%.
Am I making any sense lol?
What's wrong with say disabling llc completely, and just upping the vcore until it's stable again?, that will damage the cpu somehow?
I hope not...
From my experience it seems I needed more vcore with it disabled.
But it fluctuated less, some things acted a bit better but alot of stuff acted like my vcore was to low.
Since it was easier to get things running stable at lower voltages with it enabled I switched back.
If I were to disable LLC on my board since there is no 50% I'd be looking at 1.565v idle and 1.51v ish load for 4.3...that is 3D/game stable...LLC on I can set 1.481v and get 1.51v load...