UNOFFICIAL BIOSTAR TPower I45 M/B

[Ace-a-Rue]

the memory is a bear to set once you go beyond reasonable levels for a dual core and a Quad core, like mine with dual L2's!

i went from 470 fsb to 475 and it required a change in PL level and DDR2 Comp Reconfig...it appears that a 20 fsb jump is going to require to adjust the cpu skew control, mch skew and possibly the cpu pll...i was able to use 1.8v for cpu pll...never could do that before.

also, it appears that cpu skew and mch skew might require a 200ps difference...right now i am using 200ps for the cpu skew and 400 for mch skew.

i am trying to find the balance of memory settings and those skew controls.

it would have been nice for biostar's engineers to list all these changes in the bios in a row (on one page) like gigabyte does for the X48 boards.

[Logger]

@Ace-a-Rue: Is it just paste on the I45 or are they cemented on like other board makers are doing these days?

Hi guys first time poster. I just checked my New TPower Biostar and there is lots of gooey white stuff used on the heatsinks. Way to much goo for proper heat transfer. So the heatsinks are easy to remove. Hope that helps. I am just switching from the Abit Pro35. I have quietly been stocking Ace.

Logger

[Tony]

hmmmm..16gb is a hefty amount of memory...are you sure???...in the manual, it says maximum amount of memory is 8 GB.

EDIT:...i took another look at your image and by gosh you show 16gb of memory...how did you do that!

Its 16GB 4GB modules I even have them at 450 1:1 now for DDR900

This board needs a ton of bios work, warm booting is an issue after a bios change, psu has to go off then on and it boots, tRD and TRFC changes can force the board to non boot and actually look completely dead.

Has anyone else noticed all this?

[makaka]

The only DDR2 Comp Reconfig witch work here is number "3" . PL was set to 8-9-10-12 and 15 i can't see any deference instability remain

[shimmishim]

Its 16GB 4GB modules I even have them at 450 1:1 now for DDR900

This board needs a ton of bios work, warm booting is an issue after a bios change, psu has to go off then on and it boots, tRD and TRFC changes can force the board to non boot and actually look completely dead.

Has anyone else noticed all this?

yes tony. you are not alone!



the reboot after reboot makes me feel pain for my hard drive... kind of reminds me of my P5B!!!

[mattkosem]

the memory is a bear to set once you go beyond reasonable levels for a dual core and a Quad core, like mine with dual L2's!

i went from 470 fsb to 475 and it required a change in PL level and DDR2 Comp Reconfig...it appears that a 20 fsb jump is going to require to adjust the cpu skew control, mch skew and possibly the cpu pll...i was able to use 1.8v for cpu pll...never could do that before.

also, it appears that cpu skew and mch skew might require a 200ps difference...right now i am using 200ps for the cpu skew and 400 for mch skew.

i am trying to find the balance of memory settings and those skew controls.

it would have been nice for biostar's engineers to list all these changes in the bios in a row (on one page) like gigabyte does for the X48 boards.

I've found that, with the FSB set to 490 on my board, I'm more stable with the clock skew and driving control values on auto than with any of the suggestions that I've seen here.

This board seems to be a massive OCer, I'm very impressed so far. It's got a huge VDroop, but I've always been a fan of that anyways. I'm of the camp that keeps the power options enabled anyways, so it stays at lower voltage for 16 out of 24 hours of the day.

Does removing the heatsinks to replace the paste void the warranty on these biostar mobos? I've got a ton of Arctic Alumina sitting here from the swiftech stuff that I bought, and I wouldn't mind applying it if it'll keep my warranty fresh.

--Matt

[HousERaT]

Hi guys first time poster. I just checked my New TPower Biostar and there is lots of gooey white stuff used on the heatsinks. Way to much goo for proper heat transfer. So the heatsinks are easy to remove. Hope that helps. I am just switching from the Abit Pro35. I have quietly been stocking Ace.

Logger

Welcome to the forums......

thank you for the heads up....... Maybe some MX-2 on my nb/sb will allow me to get to 600fsb. I'm almost there but seems like the heat is holding me back. I'm pretty sure there is enough paste on my nb for three or four boards (that's usually how manufacturers spread it on there).

[Ace-a-Rue]

I am just switching from the Abit Pro35. I have quietly been stocking Ace.

Logger

...good to see you here!

[Ace-a-Rue]

Does removing the heatsinks to replace the paste void the warranty on these biostar mobos? I've got a ton of Arctic Alumina sitting here from the swiftech stuff that I bought, and I wouldn't mind applying it if it'll keep my warranty fresh.

--Matt

i presume biostar could make a case of it but if the paste isn't showing, then they would not even suspect...i doubt they would pull the heatpipe to check.

[Ace-a-Rue]

Its 16GB 4GB modules I even have them at 450 1:1 now for DDR900

This board needs a ton of bios work, warm booting is an issue after a bios change, psu has to go off then on and it boots, tRD and TRFC changes can force the board to non boot and actually look completely dead.

Has anyone else noticed all this?

what model cpu are you currently using on this board?

definitely a lot of bios work needed especially for the QX9650....my E8500 was a tad easier to manipulate the bios settings to get it stable to 520 FSB.

[Lome]

try cooling the clockgen too that seems to run toasty

[Ace-a-Rue]

i've been working overtime on getting to 500 fsb for my 9650...i am currently at 490...i finally smarten up and used the built in Memory test....i was able to find the memory settings and cpu configurations to make it pass the built-in Memtest

i just dual boot installed Vista64 Ultimate, alongside XP, to see if that OS is more tuned with this 710 bios...it installed at 490 FSB @ 8x without a hitch...once i install SP1 on it, i will install the various benches and re-run them.

[mattkosem]

I noticed that the GTL Refs with the "optimal settings" on my board are .63/.67 respectively. I'm able to run stably with a lower FSB voltage after turning them both to .67. It seems odd that they're not the same by default. Both of the other S775 mobos that I have here default to .67 on both. I'm running so far with 1.34 VID and 1.3 FSB Termination at 3.92. Maybe it's time to give 4.0 another try...

--Matt

[Ace-a-Rue]

that is different because 45nm processors were designed to work at 63%...now, that might be for the Wolfdale and the Yorkie could be different, or higher!

looks like my Yorkie, QX9650, working at 63% GTL REF, is limited to 480 FSB to function within windows...Prime 95 is going on right now at 63% GTL REF...doing nicely after 30 minutes...at 490 FSB (same multi, 8x) it would freeze or drop 1 or 2 cores out after 5 minutes.

[mattkosem]

that is different because 45nm processors were designed to work at 63%...now, that might be for the Wolfdale and the Yorkie could be different, or higher!

looks like my Yorkie, QX9650, working at 63% GTL REF, is limited to 480 FSB to function within windows...Prime 95 is going on right now at 63% GTL REF...doing nicely after 30 minutes...at 490 FSB (same multi, 8x) it would freeze or drop 1 or 2 cores out after 5 minutes.

It doesn't really make much sense to me that they would have different values by default. Why would one die be set to .63 and the other to .67? I'm able to run for 7 hours at 490 with both at .67, at which point one core fails prime. The other 3 were still running FFTs at the 10 hour mark.

--Matt

[OC4/3]

Anyone mind to prove 400+FSB Quad Core bios setting up?One of my friend is having problem over 400fsb on X3320 + Tpower.Your help will be appreciated.

[Logger]

...good to see you here!

Thank-you for the warm welcome. I pulled off the NB & Mosfet heatsink..
Both heatsinks have that crappy turdy, spongy paste especially on the NB. If you get a chance check out the architecture of the NB. Really doesn't look good for disipatting heat. I am going to fit an Evercool Vga Cooler on the Northbridge and maybe the south bridge. I have a couple of spares kicking around. THey fit perfectly.http://http://www.newegg.com/Product...20Cooler%20Kit

Hopefully I should have all my parts in house soon, so I can get my show on the road. Good Luck guys and have fun. I'am jealous.

A quick thought, I will be running Mushkin Redlines (Native speed 500) and an E8400, that will do 4.4GHz. What bios should I use? I probably want to run at 550 to 575 on a continous basis. Although 600 would be nice. Thanks.

Logger

[mattkosem]

Anyone mind to prove 400+FSB Quad Core bios setting up?One of my friend is having problem over 400fsb on X3320 + Tpower.Your help will be appreciated.

I'm able to leave everything on auto for 400 with my q9450. Once I approached 450, it became necessary to (along with the other obvious voltage changes) adjust the MCH GTL to .64 to prevent bios corruption and lack of post.

--Matt

[scorpion]

Ace:just like Logger, coming from the abit IP35pro, has any one post on E8600 yet, or would it be close to 8500. having weired core temp core 0 reads about 6c higher then core 1

[Ace-a-Rue]

It doesn't really make much sense to me that they would have different values by default. Why would one die be set to .63 and the other to .67? I'm able to run for 7 hours at 490 with both at .67, at which point one core fails prime. The other 3 were still running FFTs at the 10 hour mark.

--Matt

try 63% for both cores, that might keep Prime 95 for ever!

[Ace-a-Rue]

Ace:just like Logger, coming from the abit IP35pro, has any one post on E8600 yet, or would it be close to 8500. having weired core temp core 0 reads about 6c higher then core 1

i have my bios settings on page 1 for the E8500...it should help you get on your way...

the core temps that you read from realtemp or coretemp can be inaccurrate at the lower end of the scale...remember, those are there to throttle or shut the processor down, not to give accurate temps at the lower end.

my QX9650 have two cores about 6-7C lower that the first two cores...and..that stays the about the same when i Prime 95 it.

[Ace-a-Rue]

Thank-you for the warm welcome. I pulled off the NB & Mosfet heatsink..
Both heatsinks have that crappy turdy, spongy paste especially on the NB. If you get a chance check out the architecture of the NB. Really doesn't look good for disipatting heat. Logger

i am going to pull mine off very shortly...just finding the top end of my 9650...right now, non prime 95 is going to be 4320...it will run Hyper-PI 32M (one core) with no problem...but...my vcore is up there to hold that mhz; about 1.57v...Vdoop is not bad for a dual core (about -0.024)..i presume it will be larger for a quad like mine...i will see if i can log it before i strip the board down to check the heatpipe.

[Ace-a-Rue]

for those guys who have not "Rated" this thread, it would be greatly appreciated if you can be so kind to show 5 stars over 4 stars.

[HousERaT]

for those guys who have not "Rated" this thread, it would be greatly appreciated if you can be so kind to show 5 stars over 4 stars.

Shameless advertising.... I voted 5

[mattkosem]

try 63% for both cores, that might keep Prime 95 for ever!

I appear to need quite a bit more FSB Termination voltage to run with the same stability at .63. The results at .67 with 1.3v FSB are roughly the same as the results at .63 with 1.4v FSB. I don't pick anything noticeable up jumping to 1.4v at .67. I feel a bit more comfortable running less than 1.4.

--Matt