UNOFFICIAL BIOSTAR TPower I45 M/B

[tomati]

if you were able to get down to PL=5 on the IP35 Pro, i tip my hat to you!..the lowest i could get on the IP35 Pro was PL=7...DAMN!...you have some EXCELLENT ram!

first of all, i never said 600 FSB...i said you will get some where between 550 and 570 FSB with an E8600, E8500 or E8400.

you need to work on "Clock Gen Configuration" settings under "O.N.E." tab



it is not the AMI bios at fault; it's Biostar's Bios Engineers who did not think the voltage increments out properly...i agree...very unintelligent engineering!



REALLY!...IP35 Pro was almost a $200 board when it came out last year!...price per barrel has doubled since that time and guess what, PCB is a derviative of oil...SO...it is more expensive to build these motherboards!...make sense!



you are just SPOILED!

Sorry I forgot some detail

I tried an E8600 under the biostar , with this cpu I was able to reach 650 fsb with a friend's P5E3 premium in single channel via setfsb and boot at 620 fsb in dual so the cpu is able to catch over 600 fsb, the trouble is the biostar can't achieved this as easily as I tough, and probably there will be no point for any stability, that's what frustrating me.

For my IP35 pro , I just set 9 x 400 fsb (q6600) with ram at 1200 and memset give me PL = 5 , original trfc is 42 from ballistix and I was able to tight it to 28.

You probably right I was expecting too much.

[ddogg]

Another nudge up - Getting closer to the top spot. Dropped the 1024 WPrime over a second.

[Lome]

600+ fsb is mostly in your ram, as long as you have 8000 series cpu, but volts need to be high and mch is gonna be so loose not worth it except if you have a complex about your penis size

[tomati]

600+ fsb is mostly in your ram, as long as you have 8000 series cpu, but volts need to be high and mch is gonna be so loose not worth it except if you have a complex about your penis size

My ram can reach above 600 ( ~625 mhz with 2.35v) but I can't achieved this with the biostar and as you said , require too much volt on NB .

Anybody know what's the max safe voltage for 24/7 use with this 45µm ( right ?) NB ?

[sofarfrome]

600+ fsb is mostly in your ram, as long as you have 8000 series cpu, but volts need to be high and mch is gonna be so loose not worth it except if you have a complex about your penis size

heh-heh, hey bevis, he said penis, heh-heh

[Ace-a-Rue]

this is about it for the quad...i am about 200 mhz below my E8500.

one thing i recommend users of this board to do is re-flash the 710 bios from DOS...go here to get the bios and flashing EXE..compliments of The Rebels Haven

the best thing to do is to make your USB flash drive into a FAT32 bootable drive...here is the tool to turn your USB drive into a bootable drive...compliments of Extreme Overclocking...don't forget to also download the 98 boot files on the same page...it will need those to put them on your USB drive.

i used this flash comand: C:\AFUDOS23.EXE(space)P45AA710.BSP(space)/B /N /P (space between each switch option)

you will need to rename the AFUDOS since it has 9 alpha-numeric characters ahead of the EXE...DOS won't accept any more than 8 alpha-numeric characters.

OPTION B= PROGRAM BOOTBLOCK
OPTION N= PROGRAM NVRAM
OPTION P= PROGRAM MAIN BIOS

doing this re-flash opened up the ability to use all 4 DDR2 Recomp Configurations...before, i could only use Configuration 3...now i can use/try all 4...i found number 4 was the best and it allowed me to post to boot at PL=7...my QX9650 worked much better since flashing.

EDIT:...of course, the flashing procedure is not for those who are unsure...risks are there for mishaps...if you have a burp in your house electricity at the time you are flashing, it could render the board as inoperable...just something to think about.

MOST IMPORTANT PART OF FLASHING, DO NOT OVERCLOCK WHILE ATTEMPTING TO FLASH.

[Ace-a-Rue]

some benches to ponder

used windows XP

[rostrow]

Ace-a-Rue,

Would you be willing to run a DPC Latency test on your motherboard?
http://www.thesycon.de/deu/latency_check.shtml

thanks,
Ron

[Ace-a-Rue]

Ace-a-Rue,

Would you be willing to run a DPC Latency test on your motherboard?
http://www.thesycon.de/deu/latency_check.shtml

thanks,
Ron

this looks like it checks the driver integration with the OS..right?

[rostrow]

Right.

For some reason the test results also depend a lot on the bios. Gigabyte boards allegedly have problems with this test.

[Ace-a-Rue]

i suspect this test will probably reveal memory timings that are to tight!

i am not as satisfied with my QX9650 results...obviously the bios has problems dealing with a dual L2 cache...i see the low end 9450's doing well since they have one common or shared L2 like the dual cores...i believe my limitations are directly related to that!

[mattkosem]

I picked up a TPower i45 today to replace my failed P35 mobo. It was working fine at 3.2Ghz with my quad, so I kicked it up to 3.6(I run 3.76 normally). The system soft restarted a few times, and hasn't booted since. I've reset the bios via the jumper, no luck. After running through a few codes, I see a 63 on the LCD display, it occasionally flashes to 93. Both codes are bios related, but the board had been working fine for hours on the bios that I flashed on it as soon as I booted it. I purchased one of these boards previously a couple months ago, and the same thing happened to me. I returned it at that point, but would like to find out what the heck is happening with them. Anyone else see this happening? Surely I can't be the only person, since this has happened to me on two boards.

--Matt

[Ace-a-Rue]

I picked up a TPower i45 today to replace my failed P35 mobo. It was working fine at 3.2Ghz with my quad, so I kicked it up to 3.6(I run 3.76 normally). The system soft restarted a few times, and hasn't booted since. I've reset the bios via the jumper, no luck. After running through a few codes, I see a 63 on the LCD display, it occasionally flashes to 93. Both codes are bios related, but the board had been working fine for hours on the bios that I flashed on it as soon as I booted it. I purchased one of these boards previously a couple months ago, and the same thing happened to me. I returned it at that point, but would like to find out what the heck is happening with them. Anyone else see this happening? Surely I can't be the only person, since this has happened to me on two boards.

--Matt

always when you have a problem like this is to unplug the PSU, clear CMOS, pop the battery out and also push the board power button to unload the caps!

wait 5-10 minutes...if this does not correct the problem and get you back to a post screen, i would suspect either bad RAM or a PSU.

make sure you unplug everything from the board...even the ATX power to the board plus the CPU 12V...the PSU can hold power in its own caps for an extended period...you don't want that to feed the board...i would even yank your memory while clearing CMOS with the battery out.

when you get ready to start up, install only one stick of memory and the video card...you just want bare minimums to make it post so you can get into the bios.

if you make it into the bios, load optimal bios settings and save!

i hope this works for you.

[tomati]

Here is a rapid test for the Everest BAndwith @ 4 ghz ( 10 x 400 / ram @ 600)




I found the read value is a little bit too low ( expected in sub 10000) read is fine while copy is very weak on this board , with my q6600 on IP35 pro I have for (9 x 400 / ram 600) ~ 9800 read /

I have flashed the board after this try with the 710 bios and I will make a new test in the same condition.

I set my ram as tight I could with memset, PL5 frooze the PC, I can go on TRFC = 25 but

@ ace-a-rue, I'm impressed by your bandwith as the 3 value are very regular compared to mine.



I still has tried to hit the 600 fsb with the shared P1T1 settings found in Rebell heaven's forum but with no more succes so far.

[Ace-a-Rue]

Ace-a-Rue,

Would you be willing to run a DPC Latency test on your motherboard?
http://www.thesycon.de/deu/latency_check.shtml

thanks,
Ron

here you go!

[Ace-a-Rue]

Here is a rapid test for the Everest BAndwith @ 4 ghz ( 10 x 400 / ram @ 600)




It has a good write and read value but copy is very weak on this board

I have flashed the board after this try with the 710 bios and I will make a new test in the same condition.

I set my ram as tight I could with memset, PL5 frooze the PC, I can go on TRFC = 25 but

@ ace-a-rue, I'm impressed by your bandwith as the 3 value are very regular compared to mine.



I still has tried to hit the 600 fsb with the shared P1T1 settings found in Rebell heaven's forum but with no more succes so far.

where is "DDR2 Enhanced Mode" set to?...i have mine enabled...that could be the reason for your low "Copy" speed...also set the "NVRAM" to clear for the next bios save to get the old garbage out of there.

[tomati]

ace-a-rue,

you got the clue

this is with the DDR2 Enhanced Mode activated now



real much better

I wiill continue my test

edit : I'm able to boot at PL=6 with trfc @ 28 but I'm only @ 400 fsb while my ram is set @ 600 don't know if it's relative too the flash ? and I had cleaned the NVram as you suggested before.

[ddogg]

I picked up a TPower i45 today to replace my failed P35 mobo. It was working fine at 3.2Ghz with my quad, so I kicked it up to 3.6(I run 3.76 normally). The system soft restarted a few times, and hasn't booted since. I've reset the bios via the jumper, no luck. After running through a few codes, I see a 63 on the LCD display, it occasionally flashes to 93. Both codes are bios related, but the board had been working fine for hours on the bios that I flashed on it as soon as I booted it. I purchased one of these boards previously a couple months ago, and the same thing happened to me. I returned it at that point, but would like to find out what the heck is happening with them. Anyone else see this happening? Surely I can't be the only person, since this has happened to me on two boards.

--Matt

Matt, I'm feeling your pain - see post 164 to fix the bios, and post 196 for a setting to keep it from happening again. "Change the MCH GTL REF voltage from AUTO to 0.64xVTT ..."

[Lome]

My ram can reach above 600 ( ~625 mhz with 2.35v) but I can't achieved this with the biostar and as you said , require too much volt on NB .

Anybody know what's the max safe voltage for 24/7 use with this 45µm ( right ?) NB ?

i think p45 is 65nm
if so id say bout 1.6 is max safe 24/7

bios is very buggy, my raid card (perc 5/i) only comes up after a power down reset, the reload function doesnt work properly after an oc failure, my x1850 wont post with the 8x8 jumpers set (works on other (giga) p45), cant disable the onboard marvell chip, double option in the bios, double roms in the bios, etc.

[mattkosem]

Matt, I'm feeling your pain - see post 164 to fix the bios, and post 196 for a setting to keep it from happening again. "Change the MCH GTL REF voltage from AUTO to 0.64xVTT ..."

Wow, what a pain. That seems to have gotten it posting again. No chance to attempt the OC before work though. We'll give that a shot later.

Thanks!

-Matt

[tomati]

i think p45 is 65nm
if so id say bout 1.6 is max safe 24/7

bios is very buggy, my raid card (perc 5/i) only comes up after a power down reset, the reload function doesnt work properly after an oc failure, my x1850 wont post with the 8x8 jumpers set (works on other (giga) p45), cant disable the onboard marvell chip, double option in the bios, double roms in the bios, etc.

thanks for the tip

40GP 32M, not really fast , I don't know exactly why ? DDR2 mode and FSB tune activated, ram profil 4 + tweak via memset ( PL7 froze the PC)

12m35s 40GP 32M




bandwith @ 4ghz 8*500 / ram 625 (strap 266) + tweak via memset

10606 / 10651 / 10803 / 49.4ns



Unfortunatelly I can't get a 40GP 32M with these settings, tried with more vram ( 2.4) , loosing the ram timing ( PL and trfc ) but so far no way

Strange thing is that with PL9 I got better timing than with PL8 before it crashed ???


bandwith @ 4ghz 7.5 * 533 cas4 ram 1:1

10298 / 11290 / 11095 / 48.5ns



bandwith @ 4ghz 7*572 cas4

10700 / 12093 / 11714 / 47.6ns

[HousERaT]

anyone know if the 9800GX2 issue ever got fixed with this board?

[DiscoTech]

Another nudge up - Getting closer to the top spot. Dropped the 1024 WPrime over a second.

Very nice fsb for a quad core
Very promising board
What kind of cooling do you have on the cpu and on the motherboard?

[makaka]

Something is really wrong with this motherboard and its ram voltage , now for over 1 week i am trying to get a stable work @ 550+ but no luck , my memory and cpu are able to do 580 stable or even 600 on P965 , the OCZ ram in the 965 mobo require 2.2v for 600 mhz while in the biostar 2.3 v for 550 mhz no even stable i tried most voltage configuration but no luck , In order to boot (bios only) over 550 i require 2.4-3-2.4- and 2.5 v for 580 FSB witch is a huge .
Currently the motherboard is very stable @ 520 1:1 with 2.1 memory volt . Even +0.7 volt on the chipset do not help


work in progress ; any tips will e appreciated

[HousERaT]

I had to relax the performance level to get over 530. Try 10 or 11.