M11x clock generator ICS9LPRS387BKLF

Folks - if any of you want to improve your FPS consider a mPCIe-based x1.Opt DIY ViDock if you can access the wifi slot via the underside covers. eg: $200 GTS450 or $220 GTX460 + $20 for the PM3N card. The M11xR1 has 4500MHD onboard graphics so the x1.Opt NVidia Optimus is possible.

Would deliver better performance with a 2.07Ghz overclocked SU7300 , better yet when attach to a x1 2.0 mPCIe/expresscard port of a Series-6 chipset from the Sandby Bridge platform.

 

That would only apply if the CPU wasn't the bottleneck.

 

@Davy - Good job sir. What were your in game settings for BFBC2 to get the scores? I would literally start with AA and AF disabled and everything set to low. Follow my R1 optimization guide for BFBC2 and then try again. My honest guess is that your score on low should be about 6-8 fps higher at least than what you have now. Probably more. Also make sure you oc the gpu as well. This will help a bit.

***Edit- Never mind I just looked at your post again and you've already done what I mentioned... sorry***

Something is not right Davy. You should score higher than that. Those scores were close to my average with my optimizations. Have you edited the battlefield config file to remove the Pre-rendered frames?


Good luck and thanks for all the effort.

Best Wishes, StevenX

 

I'm running vanilla BC2 pretty much. I'll read the tweak guide now.

Update: My performance did improve and scale after applying your tweak guide. Good job, your tweaks made a 1.6Ghz Core 2 Duo perform like it's 2.07Ghz. +Rep.

1.6Ghz:
2011-01-12 17:21:44 - BFBC2Game
Frames: 4027 - Time: 99777ms - Avg: 40.360 - Min: 28 - Max: 69

2.07Ghz:
2011-01-12 17:28:24 - BFBC2Game
Frames: 5425 - Time: 99225ms - Avg: 54.674 - Min: 42 - Max: 88

 

waiting for a consolidated guide about this preferably from DavyGt...

look at those numbers an R1 with almost 60fps average in BFBC2....

and with a little push we might see an R1 finally break the 8k barrier in 3dmark06

 

He is welcome to leverage any info I've posted into a consolidated guide. I never touched a m11xR1. DavyGT has all the practicalities figured out by actually doing the mod. Seems about the second or third request for one.

 

Oh ok, I'll make the guide in a new thread later. Most of the programming information is my own anyway. I just got 1 last problem: I have no camera to take any PCB shots. Thus, I need a HQ photo of the PLL and the surrounding area to show what is being done in the mod.

One last thing: Nando4, what are the settings for the RE5 benchmarks here? I presume default settings at 1280x800?

 

Yes - RE5 benchmark edition, 1280x800, default settings. I do a DX9 and a DX10 test. RE5 gets cpu bound.

If thinking what a x1.Opt GTX460 DIY ViDock setup would gain with an overclocked [email protected] m11xR1 then consider these results:

[email protected]
[email protected] & &

I figure your CPU would be wedged somewhere b/w these two. The T6600 results include also dmcv4, FFXIV and MafiaII for more real-world comparisons.

 

ORLY?

Let's compare:

1.6Ghz U7600/5750: 44.8 FPS


1.6Ghz SU7300/GT335: 45.5 FPS


It gets better:
2.07Ghz SU7300/GT335: 52.7 FPS


2Ghz T5800/GTX 460: 49.5 FPS


2.2Ghz T6600/GTX 470: 58.4 FPS


You'll need more then some midrange cards to beat a 2.07Ghz overclocked SU7300 in these tight quarters.

EDIT: Ok, I just read that quote again and I'm feeling stupid. I thought "with" was "than". *facepalm* @.@ Back to lurking.

 

The mPCie port is port2 so can't do a x1E implementation with ATI cards either. So the GT335M would not be bettered with an external desktop solution in real-world gaming performance. The x1.Opt performance tweak requiring a more substantial CPU to improve performance, likely due to doing link-level compression.

 

That's alright with me. I don't intend to build a ViDock.
Too bad about choice being reduced though. The Optimus tweak is the only way it could improve performance?

 

This thread has got me really excited! Looking forward to a guide with images Davy. Good Work!

 

1333mhz FSB, 2.07Ghz....I'll be waiting patiently on how to do this....

 

Guys, the guide has already been made over here.

 

Thanks for the guide! I have cut the pin 14 and TME is now disabled. Problem is only that I get a full screen of green color when raising above 217Mhz. I have a Packard Bell TH36/Acer Aspire 5732Z with a GL40 chipset and 4500M graphics. Can this be related to graphics not coping with bus speed?

The stock bus speed is 200Mhz, does that mean that I can cut the 6th leg according to the schematics to get 266MHz bus speed? (0/0/0)

This CPU T6400 does 2150MHz prime stable on minimal volts (0.925), so I think it would possibly be able to handle 2660MHz on 1.15v.

What do you guys think?

 

Could be the RAM. Try taking out/swapping a module and see if you can OC further.

You want to cut Pin 2 (FS_B) for 266mhz from 200mhz.

 

Thanks for the reply!

Oddly enough the TME is enabled again after last reboot. OC wont stick now, as before mod. Perhaps the cut wasnt perfect..weird.

 

Try turning it off and on again.

Then, if it still doesn't work, try cutting the pin again.

 

Yes, did even remove battery and waited few minutes..no success. Guess I'll have to rip the thing apart again then..

I'm a bit weary if 1.15v will be sufficient for 2.66Ghz, a bit dodgy that one.

 

After a nights rest it worked again. Guess the heat expanded the pin material and it connected again. Gonne rip open anyway and do the 266MHz mod.

 

How do you know 266FSB will be stable without testing it with 1.15v and SetFSB?

 

Well, I don't. I have however done some research on HWbot on recorded OC on the same CPU family and with results there I've come to conclusion that I will have a decent chance. The best OC on stock air is 2.9GHz with 1.15v so believe it could work.
kim55's 2930.04 mhz CPU-Z run with Core 2 T6600 (2.2Ghz) @ 2930MHz
Woomack's 2904.1 mhz CPU-Z run with Core 2 T6600 (2.2Ghz) @ 2904.1MHz
SArd's 2612.5 mhz CPU-Z run with Core 2 T6600 (2.2Ghz) @ 2612.5MHz

The heatsink on Acer is a good one as well. I will use a very thin cutter, so if not working I'll simply dab on it with an soldering iron to seal it again. No biggie. Wish me luck!

 

Newsflash: I managed to cut the traces too deep and borked the motherboard completely. Oh well, there you go for beeing greedy.

 

Try some conductive ink into the hole?

 

Hmm, never occured to me, good call! I did use some solder, but I guess it didnt float down deep enough. Oh whell, it has been sent to service anyway. Waiting for cost estimate, possibly cheaper to buy new. Exciting times.

 

Good day guys. I hope you can help about my m11x r1 pinmod. I finally able to connect a resistor to pin14 and use setfsb(with code from Abo.) and confuse what is the next step to OC my M11x??? I set slide to 285 and each time i restart my laptop freeze or unable to boot properly. Is there anything that i missed??? By the way im new to SetFsb. THanks in advance.

 

it happens the same to me with same pll
my laptop is acer nav50
when i restart it freezes, it seems i have to shutdown instead of restart and use setfsb after every start
try automatically reverting original fsb before restarting with setpll

edit: once applied setpll or setfsb, pc can't be rebooted, only shutdown and start again (of course fsb changes are lost) otherwise it freeze and says something like "no device found, restart system"

once we have moved clocks sata to pll2 and src to pll3 so that cpu can be modified in pll1, those changes can't be reverted back , that's why i need to shutdown completely to avoid system freeze

 

antivirus in my laptop makes setpll lasts 30+ seconds instead of 11 seconds when i deactivate it (besides it is nod32, supposed to be the fastest)
also if i disable the following line in setpll.bat by adding :: at the begining of the line, it only takes 6 seconds to make effect:

if "%preview%"=="" call :cpu_report

disabling that line just avoids cpu report (new speed confirmation) after setpll has made effects

i have made nav50.lut (based on Alienware-m11xR1.lut) valid for acer ao532h with ics9lprs387bklf pll and cpu atom n450 1.66ghz (i need more [BCLK] values, please help):

:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
:: PLL: ics9lprs387bklf
:: SYSTEM: Acer Aspire One 532h Nav50 (166Mhz BCLK)
:: TME-unlock PLL pinmodded
:: BCLKS: 166-170Mhz
:: AUTHOR: manocao
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
::
:: Create two sections INIT, BCLK.
:: Add the (optional) PCIE section for pcie overclock as well
::
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
:: [INIT] section
:: 1. set the PLL as when setfsb does a "SET".
:: 2. Note the bytes that change when alter the BCLK, placing
:: [BCLK_B1]..[BCLKB5] in those positions.
:: 3. If creating a PCIE section then note the bytes that change when
:: alter the PCIE clock, placing [PCIE_B1]..[PCIE_B5] in those pos.
::
:: Eg: the setFSB diagnosis window OR r-w "Clock Generator" data appears as
::
:: 00 01 02 03 04 05 06 07 08 09 0A 0B 0C 0D 0E 0F
:: 10 11 12 13 14 15 16 17 [ local3 ]
:: [ local4 ] [ local5 ]
::
:: Converts to be in the following format below
::
:: set local0=0x0001020304050607
:: set local1=0x08090A0B0C0D0E0F
:: set local2=0x1011121314151617
:: set local3=..................
:: set local4=..................
:: set local5=..................
::
:: The byte(s) that change during BCLK changes is substituted with
:: the strings [BCLK_B1], [BCLK_B2] .. [BCLK_B5]. (Optional) The byte(s)
:: that change during PCIE changes is substituted with the strings
:: [PCIE_B1], [PCIE_B2] .. [PCIE_B5] from the LUT below. Ensure localx
:: is 64-bit (16 chars), pad with 00 if needed.
::
:: bytecount = number of bytes to be send to the PLL
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
:: ICS9LPRS387BKLF PLL changes to power-on defaults (BYTE:bit)
:: 0x00:b1=1 SATA=PLL2
:: :b2=1 SRC MAIN=PLL3
:: 0x0A:b2=1 CPU divider enable=enable
:: :b3=0 PCI divider enable=disable
:: :b4=1 SRC divider enable=enable
:: :b5=1 PLL2 enable=powered up
:: :b6=1 PLL3 enable=powered up
:: 0x0D:x= [BCLK_B1]
:: 0x0E:x= [BCLK_B2]
:: 0x11:x= [PCIE_B1]
:: 0x12:x= [PCIE_B2]
:: 0x15:b0=1 M/N enabled

[INIT]
set bytecount=0x16
set local0=0x678BF1EF3F108031
set local1=0x50A5651516[BCLK_B1][BCLK_B2]ED
set local2=0x3B[PCIE_B1][PCIE_B2]0924[BCLK_B3]0000

:: USING [PCIE] so need a default PCIE clock if no pcie commandline param
:: is given. Default to 100Mhz
if "%pcie%"=="" set pcie=100

::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
:: [BCLK] section. Found by placing the two PLL byte that change in setFSB
:: when move the slider noting the BCLK they correspond to. Added a third
:: byte which is the M/N enabler so as to accomodate the RST entry, setting
:: PLL back to bootup BCLK
::
::[BCLK], [BCLK_B1] (byte 0x1D), [BCLK_B2] (byte 0x1E), [BCLK_B3) (byte 0x15)
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::

[BCLK]
166, 49, 4E, 01
168, 49, 4F, 01
170, 49, 50, 01

:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
:: PCIE SECTION - move the PCIE slider in setFSB, noting the PCIE clock
:: [PCIE], [PCIE_B1] (byte 0x11), [PCIE_B2] (byte 0x12)
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
[PCIE]
100, 20, 32
105, 21, 36
110, 23, 3C
115, 21, 3B
120, 1f, 3a
125, 21, 40
127, 20, 3F
129, 20, 40
131, 20, 41
133, 20, 42

 

i found out a lot of things

when pc freezes by adding +1 to 0E, it is because pci-e bus must before be rised one step (setpll offers 100,105,110,115,120,125,127,129,131,133 mhz)

in hiren's mini xp (without hdd, booted from pendrive) i discovered all these working speeds:

fsb 0D 0E pci-e
---------------------
166.3 49 4E 100
168.4 49 4F 100
170.5 49 50 100
172.6 49 51 105
174.7 49 52 105
176.8 49 53 105
178.9 49 54 105
181.0 49 55 110
183.1 49 56 110
185.2 49 57 110
187.3 49 58 110

i stopped testing here, cpu was 1873mhz, i thought more than 10% pci-e bus may cause read-write errors

the problem came when i reinstalled the hdd, and tried all this in real xp: just +5% pci-e bus (105mhz) made the pc fail, finally stop (data loss, sata bus error)

i tried also disabling speedstep but didnt work

so, it must not be as easy as adding +1 to 0E without modifing 0D, i still need a working fsb-0D-0E list

and pci-e must also be taken into account

up to now only 170.5mhz fsb is safe (1663 to 1705, only +2.5%, not worthy)

 

do you know someone who successfully pinmodded fsb 166 > 200 in acer nav50 with ics9lprs387bklf pll by cutting fsla pin 20 and connecting it to ground through 10k resistor? pin 20 is both fsla and usb_48mhz clock output 3.3v, will usb port be disabled?
thanks

 

hi, i want to do FSLx pinmod in acer nav50 with pll ics9lprs387bklf

i would like moral hazard, nando4, davygt, dufus etc to give me an answer (i dont know how to send private messages in this forum):

what if i cut pin20? is it safe? will it affect usb bus? please see pics:

ICS PLL:


INTEL NM10 CHIPSET


MODIFICATIONS


manufacture date of my laptop is march 2010, so modification 09/03 is implemented, and what is T64 at the end of pin 20 bus?

full schematics: http://lqv77.com/download/849/

please help

 

Could anyone please post this ,pll for my D260? thanks

 

Hello!

I have an M11x R1 with ics9lprs387 and su7300. I did the pin mod, and already overclocked the cpu to 1820MHz. But I cannot use setpll, just setfsb with this code:
"C:\Program Files (x86)\SETFSB\setfsb.exe" -w05 -s280 -i10 -cg[ICS9LPRS387BKLF] -u1

I cannot get higher frequency, only 1820MHz.
And the other problem is, when I would like to use setpll and type this code:
setpll Alienware_M11X_R1.lut

THe answer is:
No .lut for this pll ... Please create one.

Could you recommend solutions for this problem.
I already did the vid mod.
But it is instable at higher frequency.

Thank You for your help!