Biostar T Power I45 BIOS thread

[thehinge1]

I am absolutely positive of the TRFC=78 setting on the I45. My P5Q-E will do the same speeds with the same kit at TRFC=55 or so, hence my thinking it is either lack of discrete DRAM skew or dirtier clocks on TPower causing the greater time needed to synchronize 2GB stix.

I know that more than 2.2v will kill them eventually. I only tried it for 5 minutes or so, just long enough to validate the lack of improvement. The label on these dimms says 2.2v, so I'm sticking with that (or the 2.168v that I get IRL). tRFC of 78 huh? I didn't need to do that on the p35, but I can give it a shot. I slapped them on my GF's P31 mobo and am still able to run the same timings that I was able to run on the BloodIron, which includes the tRFC of 46. I had to adjust the clock skew to get my FSB stable. I'm using the settings in the first page of the unofficial tpower i45 thread currently. I'd love to run 4:5 or 5:6, but my ram isn't stable at 1250 or 1300, and I'm not interested in kicking down my FSB.

--Matt

[mattkosem]

I know that more than 2.2v will kill them eventually. I only tried it for 5 minutes or so, just long enough to validate the lack of improvement. The label on these dimms says 2.2v, so I'm sticking with that (or the 2.168v that I get IRL). tRFC of 78 huh? I didn't need to do that on the p35, but I can give it a shot. I slapped them on my GF's P31 mobo and am still able to run the same timings that I was able to run on the BloodIron, which includes the tRFC of 46. I had to adjust the clock skew to get my FSB stable. I'm using the settings in the first page of the unofficial tpower i45 thread currently. I'd love to run 4:5 or 5:6, but my ram isn't stable at 1250 or 1300, and I'm not interested in kicking down my FSB.

--Matt

Still won't even post at PL8, even with tRFC set to 78. Any other ideas?

--Matt

Edit: attached current settings. I tried PL8 at 5-5-5-18 with Read to Precharge set to 5 (default) and tRFC set to 78, it was a no-go.

[Ket]

I did say that b4 was BETA and to have BIOS recovery methods handy just in case, link now gone until I figure out what the issue is.

[firebane]

I did say that b4 was BETA and to have BIOS recovery methods handy just in case, link now gone until I figure out what the issue is.

I'm not holding you responsible by any means.. I know the risks and I'll find a way back.

[Ket]

I know m8, just reiterating what I said when I posted that BIOS so there isn't a flashflood of newbs that all start ranting or something I decided the best thing to do is go back to the drawing board and start over completely. All other links are dead as the files have been removed. No fancy changes starting from scratch again, but heres beta 1.

beta 1 DL

[JWilson]

I did say that b4 was BETA and to have BIOS recovery methods handy just in case, link now gone until I figure out what the issue is.

Hmmm... it is interesting your not testing your own bios. And you do have the board!

[Ket]

I have also said not far back theres a lot of things I MUST catch up on. This BIOS work is what I do during my spare time, the rest of it I'm doing WORK. When I have time to test I will, just look in the P5Q BIOS mod thread, you will see I killed my own board to troubleshoot without giving it a second thought

[thehinge1]

I could have sworn I already posted this but it 'poofed' or I'm more tired than I thought. Only one of three P45 boards I've tested will do PL-10 at relatively low FSB, needing PL11 or 12 higher up. My board may go lower than TRFC 78 now that I think about it (hated anything under 60 for sure with my kit), but she's limping on one stick atm while I stomp on the P5Q-E/E8400 setup for a bit.

Keep in mind this was on a divider, but it's 4-2GB Rpr1066 sticks on our board: http://www.ocztechnologyforum.com/qvl/detail.php?id=29

Still won't even post at PL8, even with tRFC set to 78. Any other ideas?

--Matt

Edit: attached current settings. I tried PL8 at 5-5-5-18 with Read to Precharge set to 5 (default) and tRFC set to 78, it was a no-go.

[Ket]

I could have sworn I already posted this but it 'poofed' or I'm more tired than I thought. Only one of three P45 boards I've tested will do PL-10 at relatively low FSB, needing PL11 or 12 higher up. My board may go lower than TRFC 78 now that I think about it (hated anything under 60 for sure with my kit), but she's limping on one stick atm while I stomp on the P5Q-E/E8400 setup for a bit.

Keep in mind this was on a divider, but it's 4-2GB Rpr1066 sticks on our board: http://www.ocztechnologyforum.com/qvl/detail.php?id=29

Serious? With RCOMP config 2 and PL6 I can do 1278MHz+. You sure its not a setting you need to tweak? Take a look at my BIOS screens and try those settings

[thehinge1]

Around the time I started pushing TPower the 'new' RAM P5Q-E came in so I never pushed real hard. I did notice it pushed 1G stix easier than 2G of course. We all know how P5Q has 1:1 PL issues

I'm afraid to put the 1.25VID Eo on TPower so I've been preoccupied getting well past the 4.0 mark with it on the 'other board' whiel I wait for a tweaked BIOS.

Serious? With RCOMP config 2 and PL6 I can do 1278MHz+. You sure its not a setting you need to tweak? Take a look at my BIOS screens and try those settings

[Ket]

P5Q 1:1 stuff is alright, there is something somewhat flaky with the 400 strap from my experiences though. Anything but the auto setting (which is 1:1) while using the 400 strap caused the board to not POST, even setting 1:1 manually borked it.. very odd.

[WeldZilla]

You guys with dead boards are in great company as there are many dead Asus P5Q boards also...

You can complain about Asus and Biostar but when was the last time you loaded a new bios from either of them and it killed your board??

If you are having issues with a board any board. It is up to you to figure out just what memory it will run and run it at what ever the board will allow..

I personally feel if you want to be the great bios builder you should get the bandwidth and start your own forum as I think this is inherantly wrong

considering your board ket still lives. this tells me that these others are in fact your guinea pigs. you just make a few modifications that you think look good

then you post them and see what happens when before you posted this one you should have killed your board and never brought it up

Look around you are trying to fix one of the most incredible over clocking boards around.. For what???

I run a large Mechanic / Welding shop. Where we all know if it isn't broke we don't fix it...

Very disappointed by these threads.. Now you will jump up and say some love your bios' but, Don't for get the DEAD.

WZ

Note to all interested..... When someone like dinos comes into a thread and has doubts.... Just Stop!!!

Let the builder of the bios put it on his board... Let the same builder prove it is on his board... Let the builder prove clocks and total stability....

Then and only then should you even consider such a move. Do not be the tester unless you can afford it and knowingly want to...

Dead boards are not a good thing

[Ket]

Don't be an idiot. Again I have to point the fact out Dee reported weird stuff with an mBIOS and I flashed to it and killed my board to try and troubleshoot. In the beginning the P5Q boards were NOT good boards, and again if you look at the P5Q thread you will see my mBIOSes improve many things for many people. Lastly, for the boards that have fallen countless others have benefitted and I have always issued a warning with a BIOS I have not tested. If anyone still choses to use those particular mBIOSes, ignoring the warnings, who have they got to blame? Themselves.

[JWilson]

................... I have always issued a warning with a BIOS I have not tested. If anyone still choses to use those particular mBIOSes, ignoring the warnings, who have they got to blame? Themselves.

I'm sorry but a mod bios that is not tested SHOULD NOT be posted!

[MiniMe1978]

guys, i will get a new chip by a service in my small town 8 Euros but, the problem is still the same, why my board dont do more than 540 fsb 1:1 with my 9600 reapers? what are the settings wich give me more power? nobody say me what is rcomp? wich config is for wich result? what is fsb fine tuning? in wich combination i set for example config 2 with "timings set myself", " timings auto", pl "auto" or "myself" ? i need some facts an not results i cant get the results of somebody here if i only have one or two settings from the complete full bios... is anybody fine and send me all his biossetting screens with maybe 550 or 580 fsb and reapers 2x1gb 9600?

[mattkosem]

I could have sworn I already posted this but it 'poofed' or I'm more tired than I thought. Only one of three P45 boards I've tested will do PL-10 at relatively low FSB, needing PL11 or 12 higher up. My board may go lower than TRFC 78 now that I think about it (hated anything under 60 for sure with my kit), but she's limping on one stick atm while I stomp on the P5Q-E/E8400 setup for a bit.

Keep in mind this was on a divider, but it's 4-2GB Rpr1066 sticks on our board: http://www.ocztechnologyforum.com/qvl/detail.php?id=29

I don't want to do PL10 or higher, lol! That's a huge performance hit. I'm rock stable at 9 on mine with the same sticks that you linked with a tRFC of 46. Not sure why you're so stuck on setting it to 78, considering the performance boost that the ability to run it at 46 affords. As you'll see in the screenshot that I posted, I'm running pretty much the same settings that you linked, the only exceptions being faster tRAS and tRFC settings. I have attached my everest bench. What are you seeing in this bench, and at exactly what settings?

Serious? With RCOMP config 2 and PL6 I can do 1278MHz+. You sure its not a setting you need to tweak? Take a look at my BIOS screens and try those settings

I'll have to give RCOMP Config 2 a try after work. I've never dealt with that setting on previous boards. Maybe that's what I'm missing.

P5Q 1:1 stuff is alright, there is something somewhat flaky with the 400 strap from my experiences though. Anything but the auto setting (which is 1:1) while using the 400 strap caused the board to not POST, even setting 1:1 manually borked it.. very odd.

Other than being unable to run no lower than PL9, I've not had any issues on 1:1 with this board either. Is there some way to use the 400 strap on the i45? I only see 200, 266, and 333 on mine.

--Matt

[BababooeyHTJ]

You guys with dead boards are in great company as there are many dead Asus P5Q boards also...

You can complain about Asus and Biostar but when was the last time you loaded a new bios from either of them and it killed your board??

If you are having issues with a board any board. It is up to you to figure out just what memory it will run and run it at what ever the board will allow..

I personally feel if you want to be the great bios builder you should get the bandwidth and start your own forum as I think this is inherantly wrong

considering your board ket still lives. this tells me that these others are in fact your guinea pigs. you just make a few modifications that you think look good

then you post them and see what happens when before you posted this one you should have killed your board and never brought it up

Look around you are trying to fix one of the most incredible over clocking boards around.. For what???

I run a large Mechanic / Welding shop. Where we all know if it isn't broke we don't fix it...

Very disappointed by these threads.. Now you will jump up and say some love your bios' but, Don't for get the DEAD.

WZ

Note to all interested..... When someone like dinos comes into a thread and has doubts.... Just Stop!!!

Let the builder of the bios put it on his board... Let the same builder prove it is on his board... Let the builder prove clocks and total stability....

Then and only then should you even consider such a move. Do not be the tester unless you can afford it and knowingly want to...

Dead boards are not a good thing

This may be a good board for benching, but for high 24/7 STABLE clocks at a reasonable voltage (for 24/7 useage) this board is pretty poor. I was incredebly dissapointed by this board and I really dont like the bios. This is the most over rated product that I have ever used. I really dont like this board. I feel like I overpayed for a budget board.
1) The voltage increments are HUGE.
2) There is no way to fine tune the GTLs.
3) There are tons of standard bios options missing from the bios on this board.

[JWilson]

What standard bios items do you think are missing?

[BababooeyHTJ]

What standard bios items do you think are missing?

Real GTL options. Thats the one that gets to me. Seperate south bridge voltage adjustments. Some form of load-line calibration would be nice, but not nessesary. Memory command rate. More dividers for memory/fsb. Im sure Im forgetting something. Sorry but I really do not like this board.

[Ket]

Ditto the TPower isn't an enjoyable board, it just doesn't feel right. Plus I think the BIOS may have some real problems above 450FSB, I pushed that and it corrupted my BIOS or something, trying to figure out how to recover, can't find any info on how that BIOS-Relife is supposed to work and AFUDOS when trying to hotflash spits an error like "Error: BIOS does not support AFU".

[Praz]

I'm not sure how the board can be faulted for FSB when it can run a Q6600 at 480+.

[Ket]

Most boards can run a Q in the high 400s. I just said the BIOS may have FSB problems, I hardly said anything was conclusive

[JWilson]

Nor can it be faulted when it holds the world record for FSB and overclock! And yes those are sucide runs, but still higher then any other boards sucide runs.

[Praz]

Most boards can run a Q in the high 400s. I just said the BIOS may have FSB problems, I hardly said anything was conclusive

Guess I just don't look in the right places. I haven't seen much posted of boards running 480+ and memory at 1150 or more with a Q6600 24/7 stable.

Modding a BIOS a person really should have a way to bring the board back to life if something goes wrong.

[BababooeyHTJ]

Stability! All I care about is stability at the lowest possible voltage for the frequency. That is where this board fails. Suicide runs are useless to me. 2d benchmarks are useless to me. Volt for volt this board is a poor clocker compared to most P45's on the market.