The OFFICIAL MSI K8N Neo2-939 tips and tricks guide thread

SE my benchmark issue wasnt addressed at memory timings it was address at something MSI MAY HAVE changed with the AGP. again its a guess cuz ill swear i benched yesterday with the new graphics driver and i got around my usual 5700 in 05 and today i get 4700 and the only thing i changed was the bios.

now another thing that ahppened video wise is 3dmark is showing alot of skipping. or studdering. where as before it never did it.

I got no problem uninstalling the driver and going back to the one i had prior to this to test. Im just busy with my new toy right now :) :) :)

http://gateway.com/products/GConfig/...=m680xl&seg=hm

Hello again,

I just flashed to 1A oficial bios, all good at default settings but :mad: my system is stable with 1.8 oficial at 2.7 Ghz @ 10x270, Ram 2.5-3-3-7 and it goes to 2.8 Ghz CPu:RAM 1:1 perfectly stable @1.5 volts and max frequency is 2.95 Ghz @1.7 volts. With this 1A bios I cant get to post if I overclock to 10X270, not even at 10x250, lower I didn`t tried.
Funny thing is that if I boot at 12x200, 3x and 2.5-3-3-7, I can raise with clockgen to 2.9 Ghz @ 1,65 volts so I tried to relax my MEM timmings to 3-4-4-8 and boot at 270x10, 3x but that was a no go :( , my system cant get over post, my monitor led goes red and it stays that way until I reset CMOS.

My components are : Venice 3800+. Adata DDR600, Asetek Waterchill Antarctica, OCZ PowerStream 520 in a Stacker case.

Quote:

---

Originally Posted by 2virgule5

Just had a try of last sideffect modded bios : 150 option is not listed (only 100, 133, 166, 183 and 200), but 183 option did work well, I'm now running my X2-4200+ at 240 x 11, RAM is at 203.1 Mhz. I'll try higher HTT @ x10 later. CPU voltage (as read in bios and with CPU-Z) dropped by 0.05 to 0.1 v when I select 1.55 vcore +10% (it is at 1.55, was about 1.6 to 1.63 with 1.A3 with the same settings).

Another good news : it also seems that the cold bug disappeared, as I now can cold boot without problem at those settings (I had to reset CMOS at those with crappy 1.9 and 1.Ax).

Thanks for the work !

---

Yep because i have vapochill too i fixed the cold bug issue for myself and others. The other modded bios files even though they remove the settings for showing temp in the modbin they dont actually disable the bios from getting the temps. I disabled the bios from even trying to get them so its solved that issue nicely.

The 150 divider doesnt show because i cant add text in there. But it is selectable still. Just click enter in the memory divider screen and you will see a blank option in between the 133 and 166 dividers. Select that and the 150 divider works.

I have made a revision 2 of my bios. Im very sorry but i couldnt change Write to read delay default in this revision. The hacked code doesnt work for this setting. NFM also had this issue I will contact the maker of the code and see if the variable is wrong.

I did however force
- Max Async = 8
- Read Write Queue bypass = 16
- Bypass max = 7
in all dividers. I've left revision 1 on site to for those that dont want these memory tweaks.

Edit - removed rev 2 for the time being because i have actually forced a write to read time but its set the variable to 1 instead of 2. Will fix this issue and repost bios.

Great work SideEffect ! I don't know why MSI has professionnals developpers, nowadays the final client must program his motherboard's bios :(

@LeStat: Studdering ? hmmm, What nvidia's drivers use ? (video & gart too, plz), have you tested FEAR's demo ? It's the only game where the lags appears fastest for me ...

For Idle Cycles, I follow guides to o/c my ram on a friend's dfi's board and I havn't all tested (not enough time), but I know my sticks run @ 290mhz DC/Sync/2.5/4/3/8 - 16x/7x/8ns/256clks/5/5ns (not tested rock stable but windows boot, mini occt/streeprime okay ...)

Goodbye. I flash my motherboard again :)

edit: I wait you SE ;)

I flashed 1A rev 2 and my system died :), won`t boot at default . I went back to 1.8 official. To bad that I can`t get an official Venice bios to work with this crappy board, this board days are numbered for sure, I have enough of it.

I have tested rev1 (waiting rev2 fixed timings) ... 183 divider works okay ... I will happy for that too :) :) :) :
- Cleaned interface (some settings hidden due to unnecassery ...)
- Logos disabled & Fullscreen too
- Disabled LAN & FireWire ...
- SMART Enabled
- HTT 3x

You can Disabled Port COM no ? Parrallel no ? What you opinion about that ? (I'm order my customized bios lol)

Now just wait for really Quick Boot (no long ram tester ...)

Argh, it's 01h00 AM here ... and my hollydays are at the end ... grrr ...

Goodbye & Good night.

Hiya,

Whenever I try 250 divider is doesn't work at all.. anything above 200 doesn't work.
in cpu-z it says it's says that my ram is running at 150mhz when I use 250 divider :s

another thing, the temp sensors on the board.
yesterday it reported 41 stressed with toast and 28 idle
today it reports 34stressd and 28 idle.
I've been Priming/burning in my cpu allot so has it taken effect or is there something wrong with the temp sensors?

and whenever I hit 300 HTT in windows it instantly freezes/reboots :( I think HT frequency (instead of 3x , 2x)

xyus

Ok rev 2 is fixed and up and up for download. It modifies a few memory timings from default settings to allow higher boot speeds.

Ares323 the version you got of revision 2 had a write to read time that didnt set properly and forced it to be tighter. The bios booted fine on my pc so i didnt notice it at first. It could have caused your no boot. Sorry if it did.

The new rev2 is fully tested. Download rev1 if you dont want modified ram timings.

dividers above 200 dont work, they havent worked in any bios they have been in thats why i removed them from my bios.

Yes i can disable anything from showing in bios. Yes i also set firewire and lan off by default sry if that bothers you it can be enabled easy enough. Was kind of modding my ideal bios why i was doing this so your lucky sata isnt disabled by default :p:

Ill disable parallel port by default in m next revision but i use the serial port for my vapochill so that stays. Ill add lan back to auto in next revision to because it makes sence to have that on most people use a lan socket. Wasnt really thinking there.

hey sideeffect does ur bios support dual core chips as i just brought a x2 3800 and some gskill bh5

Its based on 1.A official so yes it should support dual core cpus no problem or lots of problems depending on how well msi added the support :p:

Quote:

---

Originally Posted by sideeffect

Ok rev 2 is fixed and up and up for download. It modifies a few memory timings from default settings to allow higher boot speeds.

Ares323 the version you got of revision 2 had a write to read time that didnt set properly and forced it to be tighter. The bios booted fine on my pc so i didnt notice it at first. It could have caused your no boot. Sorry if it did.

The new rev2 is fully tested. Download rev1 if you dont want modified ram timings.

.

---

I am mad like hell, I just cant understand why with 1.8 official I can overclock and starting with 1.9 official and all the beta`s and modded bioses, also with 1A official I can`t overclock my system , I have reinstaled my OS, I gave up the Raid 0 and went to a straight setup on Sata Drive, nothing has changed.

also i was playing around with my ga-k8ns when i went to reboot it let up like a roman candle i stood there in amazement all i had was my booster hooked up bu i want to know what does this area control. this was a refurb only 56 bucks but i was having a good time. got my bh5 to 275 all stable :) so i want to playwith it again




[IMG]http://img396.imageshack.us/img396/8340/imga00305od.jpg[/IMG]

[IMG]http://img381.imageshack.us/img381/4316/imga00283fs.jpg[/IMG]

Ares323 Post all your hardware specifications in here and maybe someone can help you.


msimax you do realise this isnt the Gigabyte forum dont you? :p:

It does support dual core. I am running one. I just flashed and I am testing the 150 divider and it is purring like a kitten. The bios looks less cluttered now. You did a great job with it. If you decide to make any other changes I have a non-priority request. It would be nice if the sata1 and 2 were disabled by default as you have to use 3 and 4 when oc'ing. Thanks for the time and effort you put into this.

MSI Neo2 Platinum 54g
Venice 3800+
1 Gb Adata DDR600
Sapphire X800Xt PE
2X Maxtor 120 GB Sata on Sata 3&4
WD 400JB on IDE 1
Pioneer 108D
Yamaha CRW f1E
SBlive
3COM ethernet card
Asetek Waterchill Antarctica CPU and VGA.
PSU OCZ PowerStream 520
CoolerMaster Stacker case.

I guess that`s all.

Quote:

---

Originally Posted by Ares323

MSI Neo2 Platinum 54g
Venice 3800+
1 Gb Adata DDR600
Sapphire X800Xt PE
2X Maxtor 120 GB Sata on Sata 3&4
WD 400JB on IDE 1
Pioneer 108D
Yamaha CRW f1E
SBlive
3COM ethernet card
Asetek Waterchill Antarctica CPU and VGA.
PSU OCZ PowerStream 520
CoolerMaster Stacker case.

I guess that`s all.

---

Ok i have a suggestion from reading the german forum and seeing peoples issues with the newer bios files. I havnt personally experianced this problem but ive heard it can happen and it may be affecting you.

Flash to the 1.A official bios but dont restart pc turn it off. Then pull out the psu cable. Move the pin to clear the cmos pull out all the IDE cables and power cables to the drives. Put the pin back in place for the cmos plug power cable back in and start pc.

It should boot the bios then. Turn off the things you dont use but leave cpu ram settings at default for the moment. save changes then turn pc off again reconect the ide cables and power plugs to the drives and start pc again.

Hopefully it will boot again. Then try and overclock.

If that doesnt work i suggest trial and error with your pci cards remove the network card and the soundcard and see what happens. Maybe you will have to use different slots. The irq for all the slots has changed in the newer bios.

If you can get to windows look in system information to see if any of the irqs are sharing a slot with another device. Sharing irqs can make some devices not work properly.

I don't know if this is possible or not but if you have a chance sideeffect could you look and see if Refresh Rate(Tref) is a setting that can be unhidden? I can set it manually usign A64 Tweaker but it would be nice if it could be set in the bios.

Quote:

---

Originally Posted by ElAguila

I don't know if this is possible or not but if you have a chance sideeffect could you look and see if Refresh Rate(Tref) is a setting that can be unhidden? I can set it manually usign A64 Tweaker but it would be nice if it could be set in the bios.

---

It cant be unhiden im afraid not yet. Hopefully MSI will add it in the next bios.

The settings that dont show in the memory options in the bios arent actually hidden they arent even in the bios code. I cant add code in there so it just uses defaults.

The exception to this is max async and read preamble they are in the bios code with all the variables and im working on a way to add them into the bios menu.

I can force Tref values the same way i forced async and bypass max. The problem is different values work better for different people in different dividers. I think its best to leave it alone for now on the bios level.

Thanks for looking. I noticed that when I use the 150 divider and then start a64 tweaker it shows the tref as 133 3.9us. I set it back to 200 3.9us. I tried using the a64 tweaker to set the setting at startup but when I do that it sets it to 100 3.9us

i just wanted to know if anyone knew what part of the board it affected or what could have caused it

Quote:

---

Originally Posted by ElAguila

Thanks for looking. I noticed that when I use the 150 divider and then start a64 tweaker it shows the tref as 133 3.9us. I set it back to 200 3.9us. I tried using the a64 tweaker to set the setting at startup but when I do that it sets it to 100 3.9us

---

133 should be looser than 200 and maybe more stable. Its actually not 3.9 us its 7.8 us athlon 64 tweaker reads that value wrongly. You cant actually set 3.9us. Setting 1.95 and 3.9 defaults to 7.8 setting 7.8us defaults to 15.6us.

With the 150 divider working I can get my system at 293x9 with the memory running at 220ht 2.5-3-3-9@1t which isn't bad for 2x1024 sticks. I need to wait for my other heatsink though to keep it at that speed.

Thats a nice overclock ElAguila. Im really quite happy with this 1.A bios i think i might keep it. Running at 270 x 11 with a 183 divider atm with tight memory timings.

Its almost identical performance to 300 x 10 with 166 divider. I lose 20 mhz of ram speed but with 2.5-2-2-8 instead of 2.5-3-3-8 i get same bandwidth in sisoft sandra and same pifast scores :)

7200 mb/s sisoft

My heatsink is fubarred right now for some reason. I am waiting for my scythe ninja to come in. It should be here by Thursday. If I can keep a handle on the heat, I have had this cpu running at 2.75ghz before.

ElAguila check out this thread it has a program that measures your tcase temperature which is to do with how well the cpu casing is applied on your cpu.

The higher the tcase the higher the overclock.

SideEffect, What chips are on your crucial ballistix ?
Prime again during the night ... argh it's too slow to testing the stability :(

++

That is an interesting thread. I ran the program and my max tcase was 63C. I was paranoid about it hitting 50C. When my new heatsink comes in I may be able to try the 2.8GHz with your latest bios.

sideeffect I sent u a PM ;)

Quote:

---

Originally Posted by FlyingHamster

sideeffect I sent u a PM ;)

---

And i answered it about 10 mins after you sent it. Was yesterday i think didnt you get it?

Quote:

---

Originally Posted by Sonic9

SideEffect, What chips are on your crucial ballistix ?
Prime again during the night ... argh it's too slow to testing the stability :(

++

---

My chips are 505 double sided 512mb sticks. I havnt taken the heat spreaders off to take a closer look.

well my stuff finaly came go start testing


[IMG]http://img398.imageshack.us/img398/3822/imga00312ar.jpg[/IMG]

a pic of modded booster a must in my book if ur using winbond with neo2 cause to much heat is in that confined area .plus with that my bh5 finally hit 275




[IMG]http://img386.imageshack.us/img386/5820/imga00330hq.jpg[/IMG]

[IMG]http://img386.imageshack.us/img386/8755/imga00347eh.jpg[/IMG]

I have a cold bug problem with my systems

MSI Neo2
Sandy 3500+ 512KB L2
Gskill TCCD 4400 LE
CoolerMaster RP450W

My system can stable at @300*10 on watercooling,

I've tried a lot of 1.36b, 1.37b, 1.9, A.1, sideeffect bios, but everytime when I set HTT over 240Mhz in bios (ex @280*10), if I turn off PC and turn on agian, it'll not boot. I have to clear bios and setting bios agian

Anyone Please help me. Thank for informations.

with every bios under 1.9 all you have to do is hit Insert then go into the bios and hit save changes

anything over 1.8 then yes you have to reset the bios.

dear god dude who modded that booster ?

i mean no offense but does that thing even run ?

what did they solder it with a torch ?

it runs

just finished the ram im in win testing now 250mgz ram



[IMG]http://img390.imageshack.us/img390/9260/imga00394ke.jpg[/IMG]

Fixes to Rev1 and 2

- Now all options available in bios will show.
- Full logo and small logo can now be enabled but are disabled by default
- Memory voltage default set to Auto
- Nvidia lan is now on by default.

Many people didnt like the hidden options so i re-enabled it all. Ill probably make a Lite Edition for those who like a cleaner looking bios with all the junk removed.

Any more suggestions on memory timings? Or shall i leave it alone now.

@ iMAX the 1.A4 is supposed to be better at dealing with the cold boot problem than the 1.A official did you try this bios to see if thats true?

Thats just what i have heard i dont experiance the cold boot issue with either versions but i did with the 1.9 bios.

LOL
No offence,, but man thats a mess :P

Is that hot glue?1?!

On that 1 pic of the booster it looks like theres a 1/4" of glue between the ram sink and the metal its mounted to? :P
If anything that HOLDS heat :P

And yeah crazy solder job :P

thanks :) :)

how do u run prime95 twice to test the cores

Quote:

---

Originally Posted by msimax

how do u run prime95 twice to test the cores

---

You have to run 2 instances of prime at a time to max both cores at once.

D3kMatrix You need to update your bios list with the 1.A official and *cough* my bios :)

Quote:

---

Originally Posted by sideeffect

D3kMatrix You need to update your bios list with the 1.A official and *cough* my bios :)

---

Yeah I know I have them all downloaded and on the server, but I need to get to bed, work at 6am =(
So tomorrow it will be updated :D

when i click on the icon twice it doesnt give me two windows

Quote:

---

Originally Posted by sideeffect

@ iMAX the 1.A4 is supposed to be better at dealing with the cold boot problem than the 1.A official did you try this bios to see if thats true?

Thats just what i have heard i dont experiance the cold boot issue with either versions but i did with the 1.9 bios.

---

Thank a lot , but where can I download 1.A4? :p:

Well you can get it from Here

Thats the Murdok version.

I just modded a version with the 183, 150 dividers but its not ready yet. Murdok is going to mod it with defaults.

Thank man, I'll try it ^_^.

this thing sucks cant do 2.6 on any voltage or any core

wonder if i should rma it and see if i get a better one or take the 15% refund hit :(

Quote:

---

Originally Posted by msimax

wonder if i should rma it and see if i get a better one or take the 15% refund hit :(

---

Try and burn the cpu in with toast it can increase the overclock. X2 processers will rarely overclock as well as X1 cpu's.

anything higher than 2.595 it just bsod then reboots now matter what volts .i have a winnie that does the samething so i dont think burn will work

Quote:

---

Originally Posted by msimax

when i click on the icon twice it doesnt give me two windows

---

Create 2 shortcuts and name them Prime0 and Prime1. Go to the properties and one one put -A0 after the prime95.exe and on the other put -A1. This tells each one to use a different core. So the target of the shortcuts would look something like this:
C:\tools\Prime95\Prime95.exe -A0
C:\tools\Prime95\Prime95.exe -A1

Launch the Prime0 and go to Options > Torture Test and Select Custom. If you only have 1GB or RAM then use 400 in the Memory to Use box(if you have 2GB RAM then select 800). Now launch the Prime1 shortcut and follow the same procedure. Now when you go to the task manager your should see both cpu's being used. I read these instructions somewhere but I cannot remember where I got them. I will post it if I remember.

Quote:

---

Originally Posted by msimax

anything higher than 2.595 it just bsod then reboots now matter what volts .i have a winnie that does the samething so i dont think burn will work

---

What voltage are you using? I am currently at 290*9 with 1.46v and sideeffects rev 2 bios. With this board I have the voltage at 1.425+5%. Also what revision are you using? My x2 3800 is the BH-E4 chip.

Quote:

---

Originally Posted by D3kMatrix

LOL
No offence,, but man thats a mess :P

Is that hot glue?1?!

On that 1 pic of the booster it looks like theres a 1/4" of glue between the ram sink and the metal its mounted to? :P
If anything that HOLDS heat :P

And yeah crazy solder job :P

---

Yeah, that looks like glue on the booster and ram. I'm sure that helps heat transfer... :rolleyes:

http://img359.imageshack.us/img359/8339/best22tb.jpg

the glue doesnt hurt :) :)

sorry for repeating this.

yesterday Speedfan reported 41 stressed with toast and 28 idle
today it reports 34stressd and 28 idle.
I've been Priming/burning in my cpu allot so has it taken effect or is there something wrong with the temp sensors?

xyus
(with computers nothing happens by it's self so that's why I'm asking)

Dear all,

I have flashed the BIOS 1.A (downloaded at MSI Global site) using WinFlash 1.76 to my MSI nForce 3 Ultra (MS-7025 revision 1). After flashing, I rebooted the machine. My computer didnt POST but a "long beep" happens.

Even I cleared the CMOS, the problem still occured. I unplugged all the power cable and waited for a while. I booted my machine up again, it can get into the window again. I have restarted the Window for serveral times and it were fine.

Anyway, I had decided to track this problem and it does still exist. Do you guys know what is my problem? Will the previous BIOS will help in this situation?

I think the furarred cpu temps are a product of the nf3 chipset in general. I had the same problem on the nf3 250 regardless of what board I used. I am currently running my 2x1024 sticks in slots 1 and 2. Has anyone determined whether slots 1 and 2 are any better or more stable at oc'ing that slots 3 and 4 or visa versa?

I only have problems with the temperatures when i dont have a cpu fan and a NB fan installed. Even though i dont use the cpu fan on my cpu just pluging a fan in there sorts it out.

I havent seen this mentioned before so just FYIish

with my UTT timings of

2-2-2-8
2-2-2-5

default to a 2T command rate, kills bandwith. Just a headsup

2-3-3-8 stays T1, I'm gonna do some more testing now, if anyone knows a fix to this issue, please let me know

What bios you using Arkangyl. It defaults to 2T with a san diego or venice E6 cpu when in bios versions before 1.9. But you have a winchester so hmm.

there are 2 bios versions made by NFM that force 1T command maybe give them a try.

I've tried 1.36 and 1.37 mod's (forget whose mod)

I'm 99% sure 1T has worked before because when borrowing some Micron 5B -G ram I had it at 3-3-3-8 265mhz and I was get 7500 mb/s bandwith which seems 1T to me.

Also Sandra says I'm running 1 channel

current setup

HTT :: 200mhz
2-2-2-5 1T
vDIMM = 3.3v rail mod
LTD = 5x
CPU = 9x
vCore = 1.5
vDIMM = auto (ought be 3.3v rail)
vAGP = 1.65
Spread Spec = off
PCI auto detect = on

Here's the Sandra output

SiSoftware Sandra

Benchmark Results
RAM Bandwidth Int Buff'd iSSE2 : 5344 MB/s
RAM Bandwidth Float Buff'd iSSE2 : 5277 MB/s

Int Buff'd iSSE2 (Integer STREAM) Results Breakdown
Assignment : 5361MB/s
Scaling : 5367MB/s
Addition : 5329MB/s
Triad : 5321MB/s
Data Item Size : 16 byte(s)
Buffering Used : Yes
Offset Displacement Used : Yes
Bandwidth Efficiency : 84% (estimated)

Float Buff'd iSSE2 (Float STREAM) Results Breakdown
Assignment : 5361MB/s
Scaling : 5366MB/s
Addition : 5210MB/s
Triad : 5171MB/s
Data Item Size : 16 byte(s)
Buffering Used : Yes
Offset Displacement Used : Yes
Bandwidth Efficiency : 82% (estimated)

Performance Test Status
Run ID : MATT64 on Wednesday, August 24, 2005 at 4:15:22 AM
Memory Used by Test : 512MB
NUMA Support : No
SMP Test : No
Total Test Threads : 1
SMT Test : No
Dynamic MP/MT Load Balance : No
Processor Affinity : P0
Page Size : 4kB
Use Large Memory Pages : No

Chipset 1
Model : Micro-Star International Co Ltd (MSI) ??? (00E1)
Front Side Bus Speed : 2x 1005MHz (2010MHz data rate)
In/Out Width : 16-bit / 16-bit
Maximum Bus Bandwidth : 8040MB/s (estimated)

Logical/Chipset 1 Memory Banks

Chipset 2
Model : Advanced Micro Devices (AMD) Athlon 64 / Opteron HyperTransport Technology Configuration
Front Side Bus Speed : 2x 1005MHz (2010MHz data rate)
In/Out Width : 16-bit / 16-bit
Maximum Bus Bandwidth : 8040MB/s (estimated)

Logical/Chipset 2 Memory Banks
Bank 0 : 512MB DDR-SDRAM 2.0-2-2-2CL 2CMD
Bank 1 : 512MB DDR-SDRAM 2.0-2-2-2CL 2CMD
Channels : 1
Bank Interleave : 2-way
Speed : 2x 200MHz (400MHz data rate)
Width : 128-bit
Maximum Memory Bus Bandwidth : 6400MB/s (estimated)

Features
(W)MMX Technology : Yes
SSE Technology : Yes
SSE2 Technology : Yes
SSE3 Technology : No
EMMX - Extended MMX Technology : Yes
HTT - Hyper-Threading Technology : No

Performance Tips
Notice 5008 : To change benchmarks, click Options.
Notice 5004 : Synthetic benchmark. May not tally with 'real-life' performance.
Notice 5006 : Only compare the results with ones obtained using the same version!
Tip 233 : Consider using the Win64 x64 version of SiSoftware Sandra.
Tip 2 : Double-click tip or press Enter while a tip is selected for more information about the tip.

Bank 0 : 512MB DDR-SDRAM 2.0-2-2-2CL 2CMD
Bank 1 : 512MB DDR-SDRAM 2.0-2-2-2CL 2CMD

When sisoft sandra reports 2cmd its not refering to 1T or 2T. If you want to see if your in 1T or 2T use athlon 64 tweaker and it will say 2t timing disabled if your in 1t.

Also you can use everest or rmma to show your timing.

Quote:

---

Originally Posted by sideeffect

And i answered it about 10 mins after you sent it. Was yesterday i think didnt you get it?

---

Strange, I never got it. :( Could u PLZ resend it?

FlyingHamster - Theres a good guide in the readme that comes with the bios files on the msi website.

sideeffect and syar -

I have now confirmed tha thte bios's you hacked is hat casued my video performance to go down the toilet.

after trying 3 different set of drivers. Sideeffects bios, Syars bios and the .36 Rev4 by NFM.... well guess what the 1.36 and 1.8 bios are the only bios's that allow my 6800 Ultra to get SOMEWHAT close to where it should be.
syar and SE bios = 4600-4800 in 3dmark 05
1.36 and 1.8 = 5700~5800 in 3dmark 05.

so maybe your not changing anything , maybe its MSI i dont know but i may get better memory overclocks but fellas i cant use this bios.

1)
i STILL cant use any add in IDE or SATA PCI card. it refuses to boot if i have a drive hooked to it.
2)
AGP bus is totally fubar'd very poor performance.
3)
and it just overall feels really slow compared to the others. maybe timings on something maybe sata rom ? whatever,,


so thanks again for you rwicked mods...

SE anychance you could take the 1.36 Rev 4 and give it the same memory timings you have for the 1.a0 bios ? the ones that give me nice clocks. ?

id like to see if THAT is whats causing it or if i can keep good agp performance and get better memory performance with your 1.0A bios timings.

It's worth a shot and maybe we can send MSI an email letting them know something isnt right with their bios.

well it looks like those gkills were holding me back i think there faulty put my trusty 5 year old bh5 im test at 2.808 at 1.47v with my x2 3800

i have up to 3.0ghz and my mb is starting to sing is that safe. i dont want a repeat of my gigabyte board

Lestat I really think those graphics card issues are either just you or quite rare. I sufferend no performance loss when going to 1.A bios.

I havnt made any changes to anything related to the AGP either so its a 1.A thing for you.

Also i havnt seen anyone else using 1.A bios complaining about graphics issues/slowdowns.

Now im not saying that the bios isnt causing the issues because it obviously is but its got to be something that doesnt effect everybody like IRQ placement, temperarture related slowdowns or bad settings.

Im using a 6800GT with the bios. 3dmark 2003 and 2005 scores are almost identical for me.

One thing i have noticed is that its slightly less aggresive on the agp bus but that means i can overclock the graphics card more. Like i was 410 mhz core on 1.8 bios and now im 425 mhz core on 1.A bios.

Hey guys, how can I increase the chipset voltage on this board? I hear that raising AGP voltage from 1.5v ups the chipset voltage.

If this is the case, how much does it raise it by? Also whats a safe limit for the AGP card over 1.5v? (I have a 6800GT 128Mb version).

Thanks

What type of ide/sata card are you trying to use. I have 2 sil680 pci ide/raid cards and they run fine for me.

Made a ballistix version of my 1.A mod bios that should work best with Crucial Ballistix. It works better than any other bios with my ballistix so i hope its a common trate.

Working on a TCCD version next with help from CD1986 and his TCCD chips.

Edit - TCCD version is out and seems to be more stable than default settings by quite a bit :toast:

Yeh, anyone with TCCD memory should try out the new 1.Amod for TCCD when it's out...works great on my system. :)

Hi,

the newest bios really sucks,

I got to 294HTT on the first bios from msi (overclocked very briefly) now I can't hit 289, 288 is rockstable but as soon as it touches 289 htt , instant freeze.

time to go to 1.A.
anyone else experiencing a problem with the official msi bios and what was helped with an unofficial bios?

xyus

i was just doin some testing in sandra with my x23800 and it showed my clock speed at 2.3 but im at 2.9 score is low also. pcmark04 shows 6500 pcmark05 shows like 1900 marks .online this score sucks hard and this is with redline at 263 and a x800xt just seems weird cpuz shows my overclock as does core center my bios is 1.a0r3..

Quote:

---

Originally Posted by Xyus89

Hi,

the newest bios really sucks,

I got to 294HTT on the first bios from msi (overclocked very briefly) now I can't hit 289, 288 is rockstable but as soon as it touches 289 htt , instant freeze.

time to go to 1.A.
anyone else experiencing a problem with the official msi bios and what was helped with an unofficial bios?

xyus

---

Dont understand what bios is the first bios and what bios you using now that cant hit what you had before.

And you say the newest bios really sucks then later you say "time to go to 1.A" 1.A is the newest bios ....

side effect the irq placements dont have anything to do with it like i initially thought.. the irq's are all the same and it still get way better performance with the older bios's.

like is iad your not touching the AGP but something within that bios is wrong. i just wish i knew what.

latencies are all the same i already checked latencies with the latency config tool.


anyways,, there is something there that is directly related to the agp performance


again any chance of having you alter the 1.36 rev 4 with the same memory timings you have for the 1.a bios ? i liked the 1.a cuz it gave me better memory clocks. Memory performance was NOT better though,, identicle settings i still got higher sandra marks ..

anyways if you get a chance and wanna play around, throw those timings on the 1.36 rev 4 and let me have it. im curious to see how it works.

:slap: 1.36 cant be modded with dividers sry lestat.

Did you try overclocking the video card Lestat? because i can overclock mine much more than before with the new bios.

mine is already overclocked .. so no i never tried..

but performance was just crap,,,lol ugh maybe ill try it this weekend and see what happens..


and i never said divider guy i said all the extra memory timings.

like opening up all the hidden memory settings and also you changed the settings that we complained about that MAY have been holding back the memory you loosened them up a little but by default i think..

anyways i dont care about the divider lol just all the memory cas timings.

anyone notice when u use the 166 div sandra and all benchs are really low when using x2 3800 with bh5 this is with 1.aor3 bios

Sideeffect, I updated my BIOS page, let me know if any of the information for you BIOS's is incorrect. =)

BTW thanks for doing all of these MODs, I'm guna have to try out the TCCD rev when I get home from work tomorrow =)

Hey guys,

Hopefully you can help me out. I recently flashed (succesfully) to sideeffect's 1.A Official revision 1 BIOS with the following string:

/py/sn/wb/cd/cp/f

I purposely left out the /cc (clear cmos) switch because my Gigaram sticks won't let me boot unless I have VDIMM at 2.85v and the settings at 2-3-2-5. After flashing, I rebooted the computer, and it POSTed. I went into the BIOS and reset everything except for the VDIMM and timings. I saved and rebooted, and system POSTed again. Next, I proceeded to power-off my system. Once I powered on, however, the system would not POST.

I tried the following:

1. Clear CMOS through jumper
2. Clear CMOS through battery removal
3. Tried working Corsair Valuselect RAM
4. Removed Soundcard, hard drive, floppy, etc.
5. Took motherboard out of case, with only video, connected

...and still nothing. I read that bad ram can destroy a system, and that's what I think I did. I tried using the two sticks of Gigaram (see signature) in another system and it would not post. Then I tried the valueselect stick, and now I have another non-posting system...

Additional symptoms include:

1. System turns on once I switch the power supply on. I don't even touch the power button on the front of the case and the system powers on. So to be articulate, all fans come on (system and cpu), hard drives power up, CD-ROMs power on and have steady power light on. If hard drive is connected, LED for HDD activity is steadily lit on the front of the case. Also, the motherboard doesn't respond to the power button when I hold it for 4-seconds or longer. It won't turn off. I have to use the same switch on the PSU to turn it off.

If any of you have suggestions, I am ready and willing to do anything, even RMA the ram. I just hope I can do something else about this motherboard. I hope it isn't dead.

with the first bios I was able to hit +295HTT with 1.9 and sideeffect his bios I can't hit 289, 288 goes perfect bus as soon as it touches 289 it instantly freezes.
my HT frequency is set to 2x & ram divider on 166 mhz so... dunno what's wrong.


and how do I install the first bios from the msi site with winflash? it doesn't contain a .BIN file?


xyus

The bios files after 1.8 often wont go as high in HTT as the the ones before. Its a common trait across all the newer bios files. My bios files should allow better memory overclocks than the official 1.A but it wont allow a higher HTT than the 1.A official because its exactly the same cpu, divider tables.

Quote:

---

Originally Posted by aphelion

If any of you have suggestions, I am ready and willing to do anything, even RMA the ram. I just hope I can do something else about this motherboard. I hope it isn't dead.

---

Problem is that The newer bios really needs a clean start. So you need to clear cmos right away as soon as you flash start it and ussually remove the power supply lead and sometimes even remove the ide leads.

Your issue is strange though sounds like a cold boot problem but that should go away after you clear the cmos and start again. I guess the memory not booting unless its 2.85 volts could be a problem.

If you have the ability to hot flash the bios chip i think that would be worth a try. Its possible the bios is corrupted and so wont boot.

If you get the board running again i can make you a newer bios that has a memory voltage of 2.85v on boot so you dont have to worry about clearing cmos.

Hi,
I found my problem.
HT frequency, if I put it to x2 I can't hit 289 put If I put it to x3 I can.

I've reached 2850mhz ;)
http://img388.imageshack.us/img388/844/2826mhz6kx.jpg

I have got higher with 200mhz but didn't take a screenie.
(when I just installed xp I got a superpi of 33.5s @ 2690mhz @ 2.5-3-3-6, now I get a superpi of 33.800 @ 2720mhz :( , time for a format )

how do you like my suicide ?
the cpu love's the vcore.
Xyus

Just tested the 3D performance of the 1.A mod (TCCD version) after seeing Lestat having problems with benchmarks.

Using exactly the same settings as with the 1.8mod bios, my 3dmark03 increased by around 150 points using the new bios, so i don't think it should be a problem for most users.

Lestat, maybe reflash the bios once more, using the TCCD version of 1.A and try 3d benchmarks again...might just have been a bad flash. I always use winflash myself without problems, but always clear the CMOS for a few minutes after flashing.

Yes i also beat my best score with 3dmark 2003 on the 1.A bios now over 14000 points.

Sideeffect: What revision would you recommend for Mushkin Redline?

hey guys been followng this postfor a while now about once a week I come here and recently iahve been on the 1.5 bios (it works for me) Well I decided to try sideffects bios to see if it would give me anything and LOW and behold no .5 multipliers in it :( I tried 1.9 and all that too. so im back at 1.5 why is there no .5 multis anymore. that is the sweet spot for me with my 3000+ venice im at 312x8.6 thats 2650 at 1.6V with my ram at 208 its corsair pc3200c2. its CH6 memory IIRC.

Also is there any chance of getting the 143mhz mem option on our boards?

If mushkin Redline is BH5 or UTT then probably revision 1 is right for you because stock timings are already tight. The revisions mostly relax timings to get higher HTT.

The .5 mulits arent in the bios files after 1.8. They didnt set a .5 multi anyway they set a 183 divider. Because the new bios files have 183 divider or they should have done you dont need the .5 multi anymore.

And no there is no 143 divider for neo2 yet.

Anyone tried the tccd bios yet? I think I will give that a try this evening. I got my scythe ninja installed and my temps are lower under load than with the xp90c so when I get home this evening I am looking to oc a little differently using the tccd bios. I am going to see how the memory does in slots 3 and 4 versus 1 and 2.

Quote:

---

Originally Posted by sideeffect

If mushkin Redline is BH5 or UTT then probably revision 1 is right for you because stock timings are already tight. The revisions mostly relax timings to get higher HTT.

The .5 mulits arent in the bios files after 1.8. They didnt set a .5 multi anyway they set a 183 divider. Because the new bios files have 183 divider or they should have done you dont need the .5 multi anymore.

And no there is no 143 divider for neo2 yet.

---

Umm come again :) RIght now im at 8.5x312 cpu-z says im at 2650mhz and ram is at 204mhz. So am i really not at 2650? or what im lost HAHAH. in more ways then one.

Yes you really are at 2650mhz and yes the ram is what cpu-z says to its just that there are no .5 multipliers on athlon 64 so its a complex series of dividers going on there.

Quote:

---

Originally Posted by sideeffect

Yes you really are at 2650mhz and yes the ram is what cpu-z says to its just that there are no .5 multipliers on athlon 64 so its a complex series of dividers going on there.

---

So if i want to stay at where im at I have to keep this bios? That really sucks. Thanks for the help...arggh.

Quote:

---

Originally Posted by CD 1986

Just tested the 3D performance of the 1.A mod (TCCD version) after seeing Lestat having problems with benchmarks.

Using exactly the same settings as with the 1.8mod bios, my 3dmark03 increased by around 150 points using the new bios, so i don't think it should be a problem for most users.

Lestat, maybe reflash the bios once more, using the TCCD version of 1.A and try 3d benchmarks again...might just have been a bad flash. I always use winflash myself without problems, but always clear the CMOS for a few minutes after flashing.

---

CD,

i initially thought that when i flashed so in actuallity i have flashed like 4-5 times hen going back and forth between syars and SE's bios. so after that many flashes nah its not a bad flash.

i dont know... just frustrating ya know..

but if SE can implement the same ram cas settings he used in 1.a that got me higher ram into the 1.,36 rev 4 bios id be happy to flash and take a spin with it. 1.36 has 4 ram timings. just open up the other ones so i can use them... plz guy :toast:

just whatever MSI did with the 1.a bios's i cant use them cuz it refuses to let me use my promise IDE controller card so no matter if i got a billion points in 3dmark i wouldnt use it i have too many hard drives that i need daily and i cant use them if i use the 1.A bios.

ggrrr MSI you need a spankin