BIOS modding for GPU OC fun and profit!!

Does this work with the new xps 15?

 

yes i m sure of it. The USB booted fine on my PC! i used a CD-ROM to install ubuntu!

 

Does adjusting Active Trip Point in unlocked bios do anything to the fan behaviour? Because as far as I can tell my fan still turns on at the same temp threshold...or am I doing anything wrong?

 

not all of the unlocked BIOS options are operational. maybe we could compile a list of working vs. non-working options, and post them up on post #1

 

Really?!? WOW so that's really useful feature to have :-D
Why are you still including this mod seeker_moc?

 

some, not all, of the options work. and in some cases of those that don't work, it may be that Windows doesn't recognise them, as seeker_moc and error-id10t post further on in this thread.

[]

 

I know, that's why i used the word "still" as now it's kinda meaningless...

BTW the bios works great apart from that, thx seeker...

 

It will eventually. The current version of NiBiTor doesn't support 4xxM or 5xxM GPUs, and I don't know how to mod Nvidia vBIOSes manually. I sent the creator of NiBiTor examples of the vBIOSes from both the R1 and R2 XPS15, so hopefully support will be coming soon.

 

seeker, stock 525m seems to suck worse than the 5730, see my main thread.
any suggestions?

Thx

 

Many of the options do work, but the only ones that are really useful, the thermal trip points, don't. Well actually, they kinda do, when you change the values in the BIOS Linux recognizes the new trip points correctly in the ACPI folder. However, the changed ACPI values do nothing. Either the CPUs internal logic, or something else in the BIOS is overriding the ACPI values. Or it might even be Windows that's doing it, I'm still testing temps vs fan in Linux to see if the values change anything.

Also, there are still a bunch of options that I haven't messed around with yet because I don't know exactly what they do. For example, under the CPU Thermal Configuration menu, there are 2 options, TM1 and TM2, which are enabled by default. In the description it says "Setting this bit enables the thermal control circuit (TCC) portion of the Thermal Monitor feature of the CPU." Perhaps disabling this will allow manually entered trip points to function correctly. If any of you are willing to try things out, please let us know the results, and I will post things as I try them out.

 

It should be, it's a much lower part. The 540M is the 5730 equivalent, and should be slightly faster than the 5730, just like the 5730 is slightly faster than the 435. The 525 is a mid-low end part, and equivalent (speed wise) to the now two generation old midrange 4670.

 

Really? the core of the 525M and 540M are identical but when I clocked the 525M to the same speed as the 540M it was still lagging compared to the 5730. Only when I clock it to 550M speeds does it beat the 5730 soundly in 3dmark06 but it still lags behind in furmark.

 

Really? Did you test all of them? Don't just test one or two options & make this assumption.

 

has anyone tried to change the color of the lights on the side of the laptop?
the lights around the "xps" on the sides, they flash red when the battery is low, but is there anyway to control them?

 

i made that remark as a "normal laptop user" and not one who tinkers with deeper BIOS settings. maybe i was as bit pre-mature there, but i don't imagine that everybody has interest in the ACPI light sensor (ALS) since, well we don't have one!

i've tried most of the "regular" settings that you may normally find in a desktop BIOS - eg. changing memory speed, fan/thermal limits, SATA settings - and if changing the setting didn't work for me, then it just crashes my computer after POST (can't boot into Windows).

at least for me, none of these settings worked - RAM speed never changed, SATA port multiplier changes ruins eSATA operations, thermal limits don't work in Windows, changing, and for some reason enabling WOL always causes Windows to take forever to shut down (literally going from 30 seconds to 5 minutes from clicking "Shut Down" to powered down).

i didn't change any of the MHC settings as i figured that i'd either kill my laptop (don't have time to worry about doing a re-install of Windows) or Windows wouldn't recognise it, and from what i've read the ACPI settings wouldn't really have any affect in Windows ("if it ain't broke...")

there are a few options that do work (toggle Hyperthreading, and virtualisation options), but i would imagine that most people would leave that as is (except for enabling VT-d for those who use virtual machines extensively). at least for me, i found no gains in battery life by disabling CPU-based options.

 

As "normal laptop user", you should be able to make proper remark. Any conclusion should based on fact. By saying none of the settings work was really pre-mature because it's clearly made without any further test. Some of the unlocked settings already tested & worked. This is not something difficult to understand.

No one ask or force you to use the unlocked BIOS. This is BIOS request thread, so request seeker_moc to prepare a "proper" BIOS for you without the unlocked settings. I'm sure seeker_moc will be happy to provide one but if not, too bad for you.

When dealing with BIOS, even the original one from manufacturer, there will be always a risk. This is common sense, anybody knows it. Any good user will know to trace back where this modified BIOS came from, to know whether the source can be trusted or not. They will found this original post (it is google-able). In the original post, I already put a warning & remind people to be careful with the unlocked settings. I already mods many BIOS (more than 100) & based on the feedback many people happy with the unlocked BIOS. Based on the number of downloads recorded, MANY people looking for unlocked BIOS & happy to use it.

If you don't have any respect towards my work, don't use the unlocked BIOS. Flash BIOS with original image & wait seeker_moc to prepare a "proper" one for you.

 

Yeap I've disabled all of this before, unfortunately it doesn't change anything. Like you said, we can see the new values in linux/ubuntu but it has no effect in Windows at least.

But as I posted earlier, Windows Power plan changes the 1st fan start value meaning I don't think we'll find a fix in BIOS.

 

first of all, I never said that I didn't appreciate any of the work you nor seeker_moc have done for us, I merely said that - with an obviously poor choice of words - that options in the unlocked BIOS don't work. In fact I'm grateful for the work you've both done. My using "none" instead of "some of", or other words to that effect, was a mistake on my part and I've gone and changed that in my previous post. I'm no forum troll (see below), and yes I used some strong words, but I never intended my words to have an iron fist affect.

some, but not all and some of these exact same settings that have been tested to be working didn't work for me. for instance, i haven't seen anyone else report back on the SATA menu, but for me the SATA port multiplier settings do not work and in fact hampered normal eSATA operations such as corrupting any file I access while Port 5's SATA port multiplier option is enabled. this BIOS-enabled option, in fact, damaged all of my virtual machines stored on my 2x external hard disks (thankfully nothing irreparable there). yes, this option "works" by allowing Windows to see multiple hard disks docked in a eSATA connected docking station, but does the BIOS option *really* work? Well, it damaged every file I accessed within Windows, some to the point beyond repair. So my answer could be "yes and no", and I can't really put a definitive answer to it; but given the outcome of my observation (and a string of profane words for good measure ) I cautiously choose to say "no".

As for "based on fact", it's kind of based on "my" fact, if you allow me to angle it like that. My observation is that most unlocked settings I tested didn't work for my 1645. Empirical fact, in only 1 book out of X number of unlocked BIOS users (the smallest possible sample size), but still empirical - I tested almost all of hte unlocked settings, except those in the ACPI and a couple other menus mostly because I didn't have the time to play with all the permutations available. Perhaps if I ran linux natively (not in a virtual machine) these BIOS options would work (as already noted by seeker_moc for some of these settings), but as a heavy/power user of Windows 7, I still stick by my (changed wording and) opinion that most of the common settings are not functional within a Windows environment.

And just so I can play the forum trolling jackass, for once in my life, in the 2 minutes it took to read this:
take a chill pill, kizwan!

 

anyone have anything to say ?

 

i can't remember which thread, either one of the "Owner's" threads, but i recall somebody said that the LED was hard-wired or something like that...

maybe somebody could crack open an Alienware application and see if it would work. then again, maybe not.

 

thanks,

was

http://forum.notebookreview.com/del...ay-stop-sidelights-blinking-during-sleep.html

the thread you were talking about ? I like the lights, just want them to stay red all the time, or something like that

 

Hey, that's my topic .
No problem anymore, because my laptop isn't in my bedroom anymore

 

yes i think that was the one... and red lights would be nice - add some festive colours during Xmas time! lol

 

lol, smart move
hmm..well yeah i don't have a problem with the lights, just that I want to control them..make the laptop more "alienware-y" lol..
i know we can't change the keyboard back-light, but apart from that i was hoping for something else..

 

Can you make one for the Dell Studio 1558 as well?

I prefer lowering the voltages to stay stable at 750/800 stock clocks. Can you do that for me please? So I can play without suddenly going into low power mode because my gpu overheats?

On Dell website, search for Dell Studio 1558 or Service Tag 3MT9DN1 you will find it fast.

 

Is there a tutorial for this? Which table should I edit to find my voltages of my gpu

 

seeker_moc and a few others spent some time explaining _almost_ exactly how to create your own BIOS, but I can't remember which posts they are from. IIRC, all these posts are in this thread, probably within the 40s and 50s posts.

 

Instead of using the hex editor, try using RBE. It doesn't work with all vBIOSes, but it does work with the 4670 and 5730. It makes it real easy to alter voltages, though I'd have to warn you and say that nobody I know of has actually flashed with a volt modded vBIOS, so you'd be trying out something that is untested.

 

bios a13 released waiting for seeker to oc it!!

 

I don't think many of us use Lojack, so I doubt it's worth re-flashing for, though any new ones I make will have it. I'll take it apart today and do some comparisons to see if anything else has changed. I'll have to rename my mods as A14 now

Edit:

I just confirmed that the microcode, SSDT, DSDT, and vBIOSes are identical between A12 and A13. I think the only thing Dell changed is what they advertised, an update to the LoJack. Unless you actually use LoJack, I don't see a need to update to A13. I will still produce all new requests with A13 though, it's no extra work and I might as well.

 

You're awesome! Thanks man..

 

I think better rename it to A12A or A13A or A12.A. This way it won't overlap with Dell's.

 

I wish I didn't update to A13 before I found this thread, I have the exact same laptop setup as you(with 4 gigs of ram and a regular hdd) and I need a little more "oompf" to play Rift(MMO) - it sits at around 20-25 fps atm and drops to 1 sometimes when a lot of things are happening. I thing the issue is more about Rift not being able to use more than 2 processors, or dxd9 and Rift not detecting my hardware.

I want to try your bios out on my laptop (using stock thermal paste and fan filter removed). Should I just use the bios flash utility with the "check for new bios" unchecked or do I have to downgrade first and then update to your A13 bios. (or can you send me or post a A13A bios).

I greatly appreciate your work.

Andrew

 

You won't have to downgrade first if you uncheck "check for new bios".

 

What modules do you modify when you unlock the BIOS? I'd like to compare the stock A12 and A13 version of those modules to make sure they're identical, and if they are, I'll just copy the ones from the already unlocked A12 BIOS to the new 'official' A13.

 

The module I mod-ed is TEMPLAT00.ROM.

 

The TEMPLAT00.ROM between A12 and A13 stock are VERY different. About 1/3 of the file has changed.... Could you be so kind as to mod the new A13 BIOS for us to unlock the hidden options again? You don't even need to rebuild the BIOS, just send me the modded .ROM. Thanks again for your support.

http://www.mediafire.com/?oyqf8jgy8fshuks

Edit: I'm sure some people would also appreciate a re-modded 1647 A11, though I think 90% of the people that use my modded BIOSes have 1645s, making them the priority.

 

Sure, I'll post the modified ROM as soon as possible.

 

Thanks.

/10char

 

Done. I has PM-ed you. In the package you'll find the modified TEMPLAT00.ROM file.

 

Hi seeker_moc,

could you make bios with 810/970 for 1645 with hd4670?

thanks

 

Sorry it took me so long, I must have missed your post. Anyways:

1645_A13.5W_4670_ 810_970.zip


All new A13.5W BIOSes contain:

Based on the stock 1645 A13 BIOS
Unlocked menus
Updated CPU microcode
Bug-fixed DSDT for Windows (A13.5L optimized for Linux available on request)
4670 vBIOS from A09, to eliminate 84C thermal throttling limit
OC'd vBIOS based on request

 

Nice work, thanks!

 

sorry i haven't been following the past few posts... so i've got a question about two of the updates:
-what is the CPU microcode; what purpose does it serve?
-what is the DSDT for Windows bug fix?

 

Hey seeker. Any benefits getting the OC clocks with the updated 1647 A11 bios?

 

The microcode is kind of like a driver update for your CPU. The one in the stock BIOS is revision 3, and the newest one released is revision 4, though I'm not certain exactly what improvements that Intel made.

The DSDT is the part of the BIOS that describes the devices and capabilities of the laptop to the OS. The Intel program that compiles the DSDT detected a bunch of errors with it. I fixed the errors because they were there, but the impact to a Windows user is quite minimal as Windows does a pretty good job of working with a buggy DSDT. Linux users will see a much bigger difference due to the fixes, though one of the fixes makes Windows see all PCI-E devices as removable. It's not really a problem, but it will fill your 'safely remove devices' menu up with a bunch of devices that you can't actually remove, hence the two different DSDT versions I made.

Not really. Unless you use LoJack, there is no reason to upgrade to A11, nothing else changed. Though I will be using it as the base for all future OC requests.

 

thanks for the info!

truthfully adding more unremovable devices to my Safely Remove Devices list wouldn't hurt me much: i have enough troubles disconnecting my external drives as it is!

 

Hi seeker, can you made custom bios for me? Based on new A13 release

Like this

150 250 0.9
150 250 0.9
150 250 0.9
150 250 0.9
150 250 0.9
150 250 0.9
750 900 1.2

1645 with 4670

Thanks

 

1645_A13.5W_DanyR.zip

 

Why do you want 150/250 on idle?