M17x R1 - Features we want in future BIOS's

If is there any way to add support for the SLI GTX 460M (or even single) will be awesome. There is no DX11 Nvidia GPU update for R1.

 

What drivers did you use when you couldn't get the 5870's in crossfire? For me, If i try to install standard ATI drivers, Xfire will not work. I have to first install the Dell drivers, and then update them via have-disk to get crossfire working with non-dell drivers.

 

That is the same for me with the 4870s but NOTHING I tried worked with the 5870s so far

Will try a couple things over the next week like using PCIe 1.1 instead of 2.0 as well

 

can you get a single Mobility HD 5870 to work at it's fully potential?

 

Hi guys,

I know you have a LOT of GPU support related questions but I am still not ready to investigate them until all of the options are unlocked. So far, the throttling issue is corrected and you will have access to all of your voltages. This is a "sweet taste" shot of one of the new menus.



Still a lot of work happening with the R2 as well, so stay tuned as I try to get the R1 and R2 BIOSes tested and ready for primetime.

Thanks,
TheWiz

 

^^
This is what we were dreaming of !!

 

I've been known to make a couple dreams come true

 

We will need a bit of an explaination about what every new option in BIOS does

 

The only confusing ones should be CPU Spread Spectrum and the FSB BSEL Override. I Need to do a little research on those myself. Tomorrow night I will be able to test the bios out! (Need to make a crisis disk first though, any links to a good guide?)

 

I'll be doing the same myself, at least for the R2 when TheWiz makes it available. Especially with the 720M and 540M. Hopefully there is a good link out there because I'd rather do some research before than after.

I think given the overclocking options, the 540M should net me some more global points on hwbot since it has hyper threading and it's a dual core. Some voltage will do me some good. I'll mostly be testing out the bios with me benching SSD. I think with the 540M being TDP unlocked, it could really be a monster. OCed FSB, and increased TDP. From what I remember, it had some headroom as far as the heat was concered as well. The 540M is the CPU I'm most looking forward to OCing, the 920XM can put out more than enough heat already. I've already validated the 540M on stock voltage at 4.1GHz and was able to bench at 4GHz. Now i'm wondering if I can push that thing to over 5GHz. Might also have to bust out the HyperX RAM which I haven't really used for any benching as well. Great time to be an alienware owner and thanks for all your work to make this possible Scook! +1

 

Remember that you need exponentially more voltage to get a linear gain in speed with processors. It get's very expensive very quick as far as voltages and heat when upping the speed. A double in speed takes 4 times the power. A triple in speed takes 9x and so on

You double the power about just going for a 40% gain in speed

These are of course not exact numbers but following the physics that governs transistors so they should put you in the right ballpark

 

Hi guys.

We have one confirmed successful flash today from a tester based on the menu shown in the last page. I will be posting full instructions for preparing a CRISIS disk and flashing the BIOS once I hear back from scooke to have two confirmed testers.

Stay tuned, should be available by the end of the day (Also, R2 users will probably see there release immediately following)

TheWiz

 

Awesome, I get back home at about 10pm ET. Currently in San Francisco. I should be able to do an hour or so of playing then before bed time (work tomorrow )

TheWiz, can you pm me a Crisis disk procedure so I can prepare that as soon as I get home. Want to make it and test it before trying your new bios. (not that I do not trust your abilities, but better safe than sorry you know)

 

Hmm, so maybe I won't quite reach 5GHz,though I was able to bench at over 4GHz on stock votalge. I doubt I'd be able to bench close to 5GHz. Just eager to see what I can do with CPU-Z. Time will tell. Do you think that the QX9300, under the revised bios, will have any more potential than it does with throttle stop and already being able to increase voltage?

 

What I hope will happen is that I will be able to boot at benchable speeds and not need ts at all. Hopefully that will take care of the memory bandwidth problem I found while benching with ts.

 

Will the custom BIOS be a better fix for throttling than TS?? TS do it great, but sometimes it can't keep the cpu clock modulation at 100%.

 

It should be as it doesn't give the CPU the chance to throttle anymore.
I think something will just quit working when the amount of power is exceeded. Still the PSU can handle a good amount above the 240W so that's a big plus.

 

Under what conditions does the R1 throttle? I'm thinking for the R2, I will probably keep tthe throttling on, mainly because I don't want to fry the CPU

 

Hi there everyone,

I have been testing out the modded BIOS for the M17x R1.

Here you have some screenies :

As you can see from the TS screenshot, there is no throttling when we have the CPU Throttle disabled via the BIOS.

I can't guarantee that there is no throttling while playing games without the need of ThrottleStop. I have been playing Crysis, Warhead and to be honest, I have seen very minor throttling. I also have to note that I have recently flashed my two 280's into 285's.

Regarding the system stability, I can say that I have not experienced a single crash. Everything has been working perfectly.

So, at least from my experience, I'm totally sure that we are almost there. One of these days we will have a perfectly running M17X R1.

Rick.

 

Impressive stuff! Is this TheWiz's customised BIOS in action?

 

Yep, it sure is TheWiz's modded BIOS.

 

It is It is

 

That doesn't make sense to me.

way back when I was playing with my T7300, I was able to get a 50% change in frequency while only needing an extra 35% increase in voltage.

Code:

 FSB  | multi | freq | volt (min stable)  | temp| fan speed (out of 1 2 3 4 5)|


[COLOR="red"](189.8 * 9), 1708.2mhz, 0.8625V, max temp = 48C (fan speed 1)[/COLOR]

(200.0 * 9), 1800mhz, 0.9000V, max temp = 61C (fan speed 1)

[COLOR="red"](189.8 * 10), 1898mhz, 0.9250V, max temp = 54C (fan speed 1)[/COLOR]

(200.0 * 10), 2000mhz, 0.9750V, max temp = 59C (fan speed 2)
(210.2 * 10), 2102mhz, 0.9875V, max temp = 61C (fan speed 2)
(220.4 * 10), 2204mhz, 1.0125V, max temp = 61C (fan speed 2)
(230.1 * 10), 2301mhz, 1.0500V, max temp = 63C (fan speed 2)
(240.3 * 10), 2403mhz, 1.0875V, max temp = 65C (fan speed 3)
(250.0 * 10), 2500mhz, 1.1250V, max temp = 67C (fan speed 3)
(260.2 * 10), 2602mhz, 1.1625V, max temp = 72C (fan speed 3)


[COLOR="Red"]*the red values show where I tried to make a frequency by lowering the FSB, but it caused stange temps[/COLOR]

EDIT: now that I think about it, I guess what you said did make sense, must have been too early in the morning for me.

 

So does anyone know how messing with the CPU spread spectrum will affect system performance? From what I gather here at wikipedia, it may cause an increase in electromagnetic interferance. I can see how this might affect the system adversly, but I'm wondering what the real world results might produce. Any idea if this could damage the system? I've read that it should be disabled for best stability and performance though, so I'm curious now. Thoughts froom those more knowledgeable than I?


@moral, I'm hoping that that holds true for the i5. I'd like to see 5GHz in a portable system. Not so much for benching, but at least for CPU-Z

 

Yeah it actually should be left disabled or the overclock might become unstable. I will remove it when I add some more jazzy options

 

Nice, sounds good. So then we will have the option to leave it as is, and then to disable it? I normally keep everything stock if I'm not overclocking.

 

Yeah, best not mess with it at this juncture, it will disappear soon enough

 

what is this?? M17x-R1 Test Bios 850-9662 - Alienware Forum - Alienware Club - Dell Community

 

That's a test BIOS for the R1 ChrisM posted.
It's a patch apparently that only replaces two blocks.No new options in BIOS.
Probably throttling Disabled but not made available as a option in the BIOS menu.
Give it a go and let's see what we get!

 

That's pretty cool! I'm likin' the work, sir. Well done!

--

And yes, there is a test patch for the present BIOS. Nothing availabe in the BIOS, but I am assuming that throttling's been majorly toned down, or disabled entirely since the system's running fine. I am glad that the R1's -finally- getting some much-needed attention.

 

Hi guys,

I have compiled a new Release Candidate for scook9 to give a try on. I have removed the CPU spread spectrum option and FSB BSEL override, and have replaced it with a much more worthwhile option.

Quoting CPUWORLD:
Intel Dynamic Acceleration technology (IDA) is a feature that boosts CPU performance when the CPU is running single-threaded applications. This is achieved by temporary increasing frequency of one CPU core when another CPU core is idle, i.e. when the second CPU core is in Deep Sleep or lower power state. How much the frequency is increased depends on front-side bus speed of the CPU:


Also, I have added the feature for Intel's Deep Power Down Technology, which is just a newer C4 power state for shutdown.



Thanks,
TheWiz

 

If suggestions are still being taken, a way to lock out stealth mode would be great. I keep having an issue with the "handshake" between the motherboard and the power brick not going through correctly and therefore it starts in stealth mode. I never use this functions and it really kills the performance, so a way to get rid of this function would be absolutely great.

 

i don't think it disabling stealthmode in BIOS(if possible) will be a solution to your problem,a solution would be to bypass the communication with the chip and use the charger as a normal one, a plus to that would be the ability to plug a "bigger" charger (e.g. 300W etc.)

 

Then I suggest your solution to be included for future BIOS rewrites. I'm sort of learning as I go on this whole thing, and never having been a very technically inclined person I'm still getting my feet wet. If you think that would be a more workable solution, then I second what you say.

 

While that would be great, it is MUCH harder than just unlocking hidden menus. Also, you would still have a hell of a time finding a larger capacity power supply that has the right voltage and plug......especially since there is only one laptop adapter ever made that I know of that is even more than the M17x's 240W (and that is the 300W of the x7200 and it uses a very different plug)

 

If that would be the only issue scook then i promise if this adapter thing will ever get "solved" i will make one for you.

 

I am fine with the 240W personally. From my testing I have found that it is happy to run as high as 280W or so. It cut off around 295W in my testing. Getting your laptop usage that high is quite a task though

 

scook, since the issue you were having with the 5870's was getting them to run in crossfire, do you think a single 6970m might be a possible upgrade for these? Do you think our heatsinks would fit? If not, do you think the R3 heatsink for the 6970m would fit in our R1's?

 

If you look at the 5870 and 6970 next to each other, you'll see that the placement of memory chips and the GPU die are nearly identical. The pictures on MXM-Upgrade.com are what I'm referencing. It looks to me like the heatsinks would be interchangeable. Therefore, all we would need is the GPU itself. If the heatsink fits, then the same fan will obviously work too.

 

If it is an MXM 3.0 board the heatsink will fit, the standard dictates the location of the gpu memory and mounting holes

The problem will be BIOS support. I was actually not even able to get a single 5870 to work

 

Have you tried flashing the card's bios? I remember someone got one Mobility HD 5870 to work in a R1. wB is the user.

http://forum.notebookreview.com/alienware-m17x/481411-m17x-r1-hd-5870-install-failed-6.html#post6857334

 

I saw that and thought about trying an older bios but did not want to bother since I knew I wanted to be using A06. Flashing the BIOS should not have been necessary.

 

what did they change between a05 and a06?

 

The touchpad polling rate is the main thing. It makes the touchpad work much better now