How to OC the M17x-R2 using SetFSB

His CPU was likely in idle mode when he took the second shot. Unfortunately setfsb doesn't manipulate the multipliers, just the bclk. Evga eleet is supposed to do that (when coupled with an extreme CPU) but for the R2 it doesn't result in any gains.

 

According to the wiki
List of Intel Core i7 microprocessors - Wikipedia, the free encyclopedia

the i7 820QM has a multiplier of 13 and the i7 840QM a multiplier of 14
the i7 920QM has a multiplier of 15 and the i7 940QM a multiplier of 16

so for the i7 820QM it is 133 x 13 = 1729 and OCed 143 x 13 = 1859
for the i7 940XM it is 133 x 16 = 2128 and OCed 143 x 16 = 2288

for the i7 820QM and mfractal's screenshot that OCed value is the same as this calculation but I don't see the same though for the i7 940XM

What am I missing here?

*EDIT*
nevermind, I figured it out so you can ignore my question
The 920XM/940X can have an extra 5% OC from the BIOS.... that's why they can achieve higher BCLK

 

I ordered 6/6, shipped on 6/16, delivered 6/18

Going to try some more PLL's today, Although if the chip is different, likely the offsets are different too.

 

Joker, what are your memory timings with the current OC?
what is your FSB : DRAM ratio?

 

We got ours around the same time, mine was delivered 6/16. So it doesn't have anything to do with how new they are or then again it could. I've got a 940xm in here so the system allows the bios to unlock the BCLK to a degree. I'm guessing setfsb is able to pick up on this and populate the registers.

 

Don't know, the M17x doesn't have adjustable memory:bclk ratios like desktop systems do. It's all very rudimentary.

 

Great job, Joker!
+1 from me.

 

So what exactly does the CPU-Z report?
By changing the bclk, do you see this ratio changing as well?

 

you're only changing the FSB. The turbo boost multiplier stays the same.

 

found a clock gen that displays the registers!


will try to mess with it tomorrow, the offsets are different.

 

ill send you $20 if you do this and it results in me being able to overclock my weak sauce i5. gogogogogogogo

 

nice....

no offer for me? *cry*


edit: I've identified ~ 10 different IC chips so far, no clock generator. There's a billion of the same sized chips everywhere. Thought it might be around the mofsetts or processor/ICH, still looking.

edit2: really not much underneath the black stuff on the bottom of the mainboard.

edit3: have the whole mainboard sans the black tape in front of me on the desktop still looking. not looking good at this point however.

 

bump for mandrake or joker if you might have any insight what this thing looks like or where it would probally be. I've found Three IC's that I think might be it, but there's really not much information on them.

Edit : well it's going back in the laptop. I've identified 3 crystal clock generators(I'm assuming for FSB QPI and PCIe), however except for the frequency at which they work, there's no information on the IC's, so today was a bust. The rest of the IC's were either for I/O or the LEDs


edit2: found it! Right before I put the black tape on I saw the magic letters: ICS (a major clock generator manufacture) writing down the numbers now, will post in a minute.
The numbers on the clock generator appear in this exact order. A quick google search did not produce any results, however I'm sure this is is (since its right next to one of the crystals)
ICS
VCS3185BL
0952
CHN
99138683

 

w000t that sounds great Lozz thank you for taking the time and ripping your system apart
This was the one moo gave me, and the one I have the datasheet for, which didnt work with abo >> SLG8SP585V

 

really? Well that's depressing. At least I learned how to remove my mainboard.

 

Its the same one?
Well in that case maybe we can allow abo to just support it, so we can populate everything. A bit of a hassle but is it worth it
If you can get abo to do that I'll send you the datasheet as its completely meaningless to me

 

I didn't see SLG8SP585V anywhere on the IC chip, so would say it's not. I thought you were saying the numbers I posted were the same that moo gave you.

edit: Pictures!


yeah, my camera sucks. The Clock generator is the square IC, right above the right side of the PM55 heatsink, next to the clock crystal that says 14.xxxK


edit: if anyone uses these, they *do* make contact, so keep that in mind.

 

Very sexy Lozz, thanks for the pics Do you mind emailing abo with the clock gen?

 

I'll email, I just hope he can confirm that he can get it to work without me wasting $6 and him tell me the same stuff 'opps, nope I can't, but thanks for the money!'

 

Seeing as its a different clock gen, it will most likely work. He didnt ask me for money even when he got the datasheet he said he cant support it, I tried telling him I'd pay extra, he sent an email with bold and red fonted letters "I can't support it" Man that datasheet was a loss it cost me 40+ wiring fees...

 

Well I'm glad I didn't pay him...

 

meaning:
hey abo, how ya doing. fine i hope.
say, i have this pll, can you help me?
thanks.

speculation of course:
because that same thing happen to me...lol

 

I expect some more professionalism from somone that wants to charge for their software instead of snobing at potentional customers because they don't say 'pretty please with a cherry on top'. all I asked was if it would work or not, not, 'can you code this mak it work and update your freeware?'

 

that's your opinion...
those rules (quotes/sayings/feelings on the matter) apply to you, not the country he is from...(japan)
just an observation....

 

which is why I'm typing up a kiss-and make up email now. just realized it was a .jp address.

 

LOL, was it that bad Lozz
Man, he really got pissed

 

I don't think it was that bad, nevertheless

 

Haha well hopefully abo accepts the apology and can do something. You should mention to him that some of us are using the other PLL and it works so that it is possible. He may just tell you it's not possible because Dell uses some sort of TME lock (I think he mentioned something like that before, can't remember).


Edit: Check out this link, it's quite informative: http://forum.notebookreview.com/har...verclocking-methods-examples.html#post5002684

 

he did, he doesn't think the numbers I gave him are a clock generator and wants me to open it back up. even though it was right next to the correct crystal( which is the 14.xxx one) I'm on the phone right now, so I'll update whn I get home.


ps. he called me Andrew-San hehe

 

reminds me of dragon ball z.

 

This is what he asked Joe in an e-mail:

1. About the clock generator.
If it is possible, Could you examine the motherboard? (Only this method exists.)
IC near quartz vibrator(14.318MHz, Argent metal casing)

 

LOL is it weird that was exactly wha I was thinking? Vegeta ftw!

that's what he asked me, among some other things.

 

mad at the world vegeta!..lol
watching dbz now..lol. while waiting to see if you get this over clock going.

 

Yeah he called me Quadzilla-san when he was helping me with the W90...

Seemed like a good guy back then but i think things have changed since then...

 

He probably got tired of people demanding help from him and once he gave it, little thanks in return. Plus I'm sure he's got a lot going on in RL.

 

Yeah i know he ceased all development of Setfsb for sometime because of the demands and abuse from people....

I cant blame him one bit for charging for this product ..

 

If Abo can crack this PLL I'll gladly send him $10.

 

Hey maybe abo can use what you discovered into setFSB that way it will be less of a hassle for you

 

I'm guessing loz told him about it already..or should have anyway.

 

yup, I explained that other people were able to at least change the offsets.

 

i have emails back and forth with Abo as well, with no luck. +1 to lozz for taking this thing apart, i dont even want to put the 6300 wifi in =). ill throw $20 at whoever unlocks this pos as well. i need to get more power out of this Cmon!

 

I just sent him what I could, I don't expect he'll be able to do anything with it, I'll have to tear the laptop down again to verify the ICS chip (since it doesn't seem to exist in any form on the internet) or call up ICS for the datasheet. It seems alienware is using super-secret Area-51 clock generators. Or my eyesight is just not what it was (22years is old).

 

Lozz if you've got a good camera, try taking close up pics of all the chips and then we can figure out which one it is from there. I'm hoping we get this cracked because I've got yet another system replacement coming. There's dirt smudges underneath my E2E panel and I asked for a user replaceable panel replacement but the rep. decided it would be better if I had a new system outright. Anyway, if the new one has that crap PLL that nobody can OC, I'll just swap the LCD's and call it a day.

 

I'm taking it apart again as I type. I'll try to figure out the manual focus in my digital camera, but I'm not sure that it can focus on objects that close. I'll give it another twirl for sure however.

edit: my camera simply does not have enough focus or resolution in order to clearly take a picture with the lettering on the IC chips, my dad has a *much* better camera the City gave him that I'll ask if we can use, but I doubt this and it won't be tonight if we can. The lettering is very very very small as you know joker. I have to use a magnifying glass and have 20/30 vision or so even. It also seems I was not incorrect about the clock generator numbers or even which one it is. There are 3 crystals(14,24 and 32khz crystals), and the only one I can see that could be it is the one I wrote numbers for, the other is for the IEEE controller and another is for the some other controller, although I doubt it is this one since it only has 6 leads. The *only* other IC controller I have identified as a candidate is by the 24.xxhz crystal, and that has the following numbers on it:
SST25VF0328
66-41-S2AF
094029 There may be more numbers there, but it has a silver drop of paint on the IC, probably from assembly.

That was the 6 lead Controller, and I think it's just for the LED's or some other controller.


edit2: if there's anything else you want me to look at, let me know real quick before I put the magnesium plate back on fully.

edit3: moar picture!(s)




edit4: the RICOH controller is for the SD card and not the IEEE. I think It just works off an IEEE interface, this makes sense since the I/O cable connector is right next to it as you can see in the picture.

 

just stress how much money is in it for developing the correct code for the pll

alot of M17x's

 

attention:You're going to scream when you find out what the problem was


hint: It can be fixed in windows, with 1 mouse click.

 

ROTFLMAO And what did you do to get it fixed?

 

Step 1) change windows font resolution from medium or large to normal(in the control panel under font sizes) .

Step 2) adjust registers because SetFSB can't reproduce large font sizes!!!

Abo told me to lower the font size in SetFSB when the diagnostic data would not come up. So I did that.. and BAM!

I seriously screamed like a little girl for a second. When I found out it was so obvious and easy..

 

w000t Its working Thanks Lozz got it to 150 and im trying to run benchmarks now. Hopefully they improve

 

LOL I quit along time ago but i guess this might help

3dmark06 ofcourse