M17xR2 Bios A09 Discussion Thread

I had the same problem as you after flashing new vbios and drivers and bios.

What I checked was if it runs perfectly in safe mode. It did.

So I did a system restore, and it fixed the issue.

Some error corrupted win7.

 

Hey

It says my BIOS is unflashable. I get this message when I try to flash from A04 to A09:

Part ID of system BIOS and new image File are different (system = 98Q file = 35k)
The system platform may be different from the Image file. Are you sure you want to proceed?

If I proceed I get an Error -144 and it says it's unflashable. I bought the laptop refurbished from Ebay, thanks.

 

I reformatted. Issue gone. The only other thing I noticed(and I haven't checked since reformat) is that the GPU fans speed up with the CPU fan. I ran OCCT for the CPU and the GPUs fans pick up at the same time. I will run again soon and post some screenies of HWMonitor.

 

kidio007,

It may sound stupid, but make sure you know that this flash is for the M17xR2and not the M17xR1. You don't have a signature line so I don't know your specs. But the A09 is for the M17xR2 and the A05(discussed in another thread) is for the M17xR1. If you have an M17xR2 then that is a strange problem. I did notice in another thread where you talk about loud fan noise that you did say that you had a q9000 cpu, if this is the system that you are talking about then you have an M17xR1 system. The A09 bios is for the M17xR2 system.

 

Yeah I have a R1 then
Thanks for clearing that one up, I wonder how nooby I can be on these forums, not the first time I do mistakes like these

 

Hmmmmm. Here's the screen schot of the OCCT after I ran it. Notice the max temps on the GPUs. Then notice the max speeds on the GPU fans. Why are they going so fast? They're not evem warm, much less hot.

Now, I don't know if it was this way with AO8 or not. Didn't really have it long enough to check before flashing to AO9. Anyone????

 

I wouldn't trust the fan speed reading too much.

 

fan speed is quite accurate actually.
fans are ALWAYS rotating, even if the temps are 20 degrees fans will be around 1800RPM which is more or less what i see in the screenshot.
This behavior is exactly the same with A05,A07,A08 and A09 (those i tried, but i am quite sure that those BIOSes that i haven't behave the same)

 

The reason I say I wouldn't trust the readings is that I have never seen them change. I am talking about the "GPU Fan Speed" reading listed under each GPU at the bottom of the screen, which stays at 30% no matter what happens. Even when I can audibly hear the fans spin up those numbers do not budge.

The actual RPM readings on the other hand seem to vary during usage, so they might be more reliable. The percentage readings are definitely not correct though.

Also, GPU-Z never succeeds in displaying fan speeds, so my assumption is that the sensor readings are not able to be interpreted by the software correctly.

 

Yeah, those 30% readings is what the video card reports which is nothing in our case.
I helped Mumak (author of Hwinfo) a little to implement the code that displays fan speed in rpm and it is very specific to the m17x-r2.
To be able to read it Mumak reverse engineered our bios and saw how the bios reads those temps from the EC.
And since the EC is what controls the fans I guess it's pretty precise

 

I believe the readings. What I'm trying to point out is that the GPU fans speed up to max with the CPU fan without ANY temps increase. Notice the low and max GPU temps, but the fans show full speed. And I did reset the HWinfo readings before I ran OCCT. That's what I'm screaming. The GPU fans should not speed if just bc the CPU fan does. Right??? Is this also an issue with other BIOS's?

I edited the screenie.

 

Do they really speed up, or is this just from what you read there, your CPU fan should be well turning at those CPU temps, and should be the main noisemaker in my oppinion at those mid-high temps. The GPU's should be on their lowest speed at those temps (level2 >50°C).

From what i read, i would say the GPU speed is inaccurate in that application, like in most other monitoring apps awell, even in GPU-Z. Your picture even shows 30% max for both GPU's which would be the same as in GPU-Z and that would be "correct". Don't trust those monitoring apps too much.

If you really have 3 fans turning at max. then your noteboook would have already taken off by its thrust

Please check the noise level, with and without running furmark multi-gpu, which will heat up your cards decently and you should hear the difference drastically! Cheers.

 

no no, the gpu fans speed up with CPU temp increase. even if the gpus have 0 load.

 

that's right. and I find that quite unnecessary. the fan control algorithms should be more clever.

 

mfractal

Your correct

 

Oh, they do indeed, now that's new to me indeed.

Either i'm not noticing this or i'm not correctly evaluating this issue.

So from what you guys say, and to test this...

I should get the same GPU fan usage as when i run a CPU Stress test (hitting high temps on CPU) and a GPU stress test seperatly (heating up GPU's), i should get the same result (ignoring the CPU fan for the moment)... Correct?

 

that's correct. It seems that all three fans are triggered together regardless of whether the heat is on the CPU or the GPU(s) or both

 

I can confirm this, just tested it via IntelBurnTest and FurMark, and it's indeed the GPU fans comming along the CPU-Fan and vice-versa. That doesn't make sense to me for some reason, as the GPU'-s are well isolated from the CPU, no? The GPU's did not even heat up slightly whenever the CPU was under load previously and imo that issue should be eradicated. What alex already mentioned, that's a weird algorithm

Nice find guys. Sometimes even i don't notice something until someone tells me

 

i actually stumbled upon it and upon the fact that the fans rotate even when everything is super cold when i was helping Mumak debugging the EC code readings.
I really don't mind though, the cooler the laptop is the better

 

Yeah, sure sure, but when there is no direct influence from produced heat on the GPU's to the CPU and the opposite, why can't it be indivdually, i mean you have thermal sensors for each component.

Is it a problem of complexity to write some monitoring/regulating porgram into the bios or wherever this info has to be written to? I'm aware that the less complex a BIOS is, the less issues you should get with it.

Did Dell forget to look into its M17x behaviour list and forgot to add the thermal regulation or did they change that on purpose and see how many people notice that?

 

Well what controls the fans is the BIOS and to be able to modify that behavior we either need to ask Dell nicely (but i'd say it's a very low priority fix) or reverse engineer the BIOS and modify the code which is EXTREMELY difficult, not to mention dangerous.
Dell actually didn't change that, it's like that since the beginning i believe.

I wouldn't worry about that too much though, the only downside i can see to that is more noise from fans rotating fast, and fans wear.

 

Well that certainly explains why pmassey was experiencing what he saw - one of the CPU cores hit 80C which is going to rev up those fans pretty hard.

This also means that anyone considering doing one of the various cooling mods should probably do them all while they are in there, unless fan noise isn't a consideration.

 

No i'm sure if they did that in an actual point of presence, there is a reason for that.

I agree, that it's not a priority, but when you say it will have some influence on the bearing and wear of the fans, always going along the CPU fan, just because they are "forced" to, where it would be not necessary, then it's not very acceptable for me nor should it be for dell.

So i'm asking nicely now as you suggested: Please Dell, if you can fix this issue with minimal effort, please change it, so the fans can act individually on their corresponding component, ahoy

 

Soooooo (patting self on back), I was right?!?!?! And it's not a big deal really. Just didn't know if it was normal and was in previous BIOSes or just this one. Woooo hooo. I give myself rep!!!!!!!+1+1+1+1+1. JK. Nice to know I;m not alone.

 

The reason all the fans rev up is because the heat builds up inside the case and needs to be vented quickly.

 

correct me if im wrong, but arent the fans specifically designed to cool only the specific parts, ie. cpu and gpu? the fan air intake is only channeled through the heatsinks so logically, it wont really affect the internal case temperature because there isnt any airflow running through the entire internal case. fan sucks in cool air, direct them start toward the heatsinks and exit as exhaust hot air. that's it. it didnt run through the case before being vented out. and the fan case is designed so that very little air will escape and are all forced toward a determined path, which is through the heatsink and out.

so again, if the cpu is under load, and the gpu arent, there is still no reason to rev up gpu fan. that could dramatically lower the noise level of the system. what do you guys think?

 

5150Joker
Off Topic but what you buying next after you sell your R2?.

 

You can never have enough cooling IMHO i am deaf as a post anyway. Fan noise couldn't care less

 

OT

@Joker. Lemme get those 5870s cheap and I'll send you my 4870s!!!!!!

Back on, noise isn't an issue too much. Just why run the fans if it's not needed? Just gonna burn em up faster.

 

That's what 3 year CC is there for

 

i agree. it's about both noise and giving the fans a longer life span. i dont mind the noise too, because when im gaming, the sound is not much compared to the sound from speakers.

however, when not gaming or just doing some cpu intensive tasks, it'll be good to have lower fan noise if given the choice. the unneccessary loud fan noise can sometimes disrupt work.

though not a really huge issue, if Brian and his engineering team can come up with a smarter fan control algorithm, that is definately a big bonus for the alienware m17x community.

 

I may be in a position to have to use Dell's CC. I noticed while playing a game that the usual fps of 50 would suddenly dive to like 22~24, hover there for about 40 seconds then come back up. Once it starts it's cadence is like every minute or two this happens. I have tried flashing up and down the versions, now on A09, same thing. I believe my problem just started in the last month or so.

HWInfo shows my GPU1 fan never getting past around 1630 for fan speed, while the GPU2 fan is at 1900. This is while just at the desktop, no game. I suspect I am getting heat load, causing the CPU and/or GPU1 to throttle down.

I know if I touch the top left and right of the case the left gets pretty hot, but the right stays relatively cool, this while gaming on a Cryo notebook cooler btw.

Anyone have any suggestions on what I should try next or does it sound pretty conclusive?

EDIT:

Running STO with pretty much on (yes, I know it sucks) my fan speeds read thusly:

CPU 3660
GPU1 1887 <--------- ?
GPU2 4302

Agh. All I can say is AGHHH!

Hey Sauce, I guess the tack here is that you have to claim faulty 4870's to swap to the 5870s?

 

Docsteel

That's strange that your fans are not running similar RPM'S mine regardless what i am doing all seem to be pretty close.I am in no rush to get 5870's to be honest 4870's do the business for me while remaining cooler.But i guess i will probably get them given to me via tech support in the future if my 4870's decide to die.

 

Hey Docsteel,

Yup I'd say theres something wrong with the pads or thermal paste on that GPU. I had some screwed up temps and fan speeds as well and after I reapplied thermal paste and put in new pads all is good now.

If its not the thermal pads or paste it could be the heatsink or the GPU fan itself. Best to get it changed, ring up dell.

These are my temps during normal usage. Everything but without gaming. Running videos, a heap of other tasks and applications.

Also my fan speeds used to hit the 4k rpm as well but now after I fixed up the thermal paste and pads it only hits a max of 3300rpm max fan speed which is loud but no where near as loud as it used to be. Temps are also much much lower. The max fan speeds were only hit during gaming or benchmarking.

How are the temps on the GPU 1? really high since the fan speed is quite low? Otherwise it looks like the GPU is not being used properly. All the load seems to be on GPU2. Hows the performance in games and benchmarks? Any problems?

Cheers.

 

So there is a fan issue? Here I thought I was just imagining things and maybe never noticed them before. Thought it was sort of odd that just chilling on forums and stuff my GPU fans would kick in for a few from time to time now, even with a cooling pad and a relatively cool room (67*F). I've not checked temps ever but I would assume they are nowhere near close to the temps they should be for as loud as they are getting, use to only notice them at these levels when gaming.

 

You must have some dust buildup. Fans shouldn't kick in while browsing the net especially on a cooler unless you run some CPU intensive stuff in the background.

 

One point to consider, if in fact the fans were left to their own devices (thermal sensor) and did not kick in together, would we have the same level of cooling? meaning, would we see a rise in temps? follow?

Now, I'm not saying this is the case nor have I looked at any of the info you all tracked down - it was just a random thought I had as I am catching up on the thread and noticed the thermal q.

 

It's good thought, i think only dell knows how the heat is spread when the fans act individually and if the heat produced by one component will influence another component, if the fan would not already spin there and how the airflow inside will be extracted outside.

Might be worth a test for dell to see how that works out (maybe they already did), i just did not notice that it was always all fans together, kinda surprised as i got told

 

from a logical point of view, having each fan controlled by each component's thermal sensor is the best and most efficient way. it doesnt matter how heat is generated or distributed, once the sensor detects a sudden rise in temperature, the corresponding fan will kick in to cool that component down. again, the internal case temperature (not the cpu, gpu obviously) has very little to do with how fast or slow the fans spin because air clearly is not routed throughout the internal case before vented out. air is only directed toward the heatsink and no where else. and to back this up, each component is well isolated from one another. in other words, they are well spaced in between so for ie. if the cpu heats up, it wont directly influence gpu temp and vice versa. u can run wPrime for hours and not see any rise in gpu temperature.

my best guess is, at the very low level control system, the BIOS is not able to read temperature feedback from gpu sensor. that is why it automatically assumes that if the cpu is hot, the gpu will be hot too so all 3 fans kick in at the same time just to be safe. then again, it's best to hear it directly from Brian and his engineering team.

 

Just a note to all, Brian sent me a heads-up that he is on vacation. Do not expect to hear from him before the 4th (October).

 

oh my.. what will we do????


DOOOM DOOOOOOM... lol

Well hope he is enjoying it. Thanks Batboy

 

thanks for the headsup BatBoy

 

Are you certain you have the M17x-R2 and not the M17x-R1? The M17x-R2 is the one with the Core i CPU, the M17x-R1 the one with the Core 2 CPU. This BIOS is only for the M17x-R2.

 

The man deserves it for putting up with us.

 

my i7 820 will not hit 3.06 in turbo mode benching 1 core anymore. It hits 2.93 and holds there using prime 95 and hyper-pi. I guess this is the BIOS a09 doing this right? Is it possible to flash back to A05? (i think thats the version before A08)

Please help as this slows down some rendering I do a little bit.

thanks
AwHw

 

for the 1st let me introduce myself and sorry if my english bad
i'm rizki from indonesia. i'm new user of alienware m17x-r2 product.

i had problem after flashing bios to A09
my Win7 system can't detect, so it can boot to win7 now.
i have try to diagnostic via DVD resource And bios diagnostic.. but everything running well.
in bios harddisk detect 320x2GB but when i want to reinstall it, Hardisk seting on win7 instalation only detetct just 1 Harddisk and it can't reinstall.

please help me guru now my m17x cannot do anything.. only can boot till bios

 

You said you've been into the BIOS but while you were in there did you happen to see if you needed to reset your RAID settings to the way they were before you flashed to A09? This sounds like the problem to me anyways.

 

Hi Rizki,
Go into the system BIOS, check your HDD setting and ensure it's in the correct mode (RAID or AHCI).

 

What he said flashed it yesterday had the same problem it was my raid settings in bios.

 

No it is not A09's doing Intel doesn't guarantee that all Core i7 processors will hit the Max turbo boost. Increasing Performance with Intel® Turbo Boost Technology

read the little foot note on the bottom Also the Turbo monitor gadget doesnt always read right As in Turbo Gadget it will show my overclocked 720 at 3.16ghz, but it is running at 3.25 ghz really in CPU-Z