1.8GHz Pentium M overclocked to 2.4GHz

Associate
Joined
30 Jun 2008
Posts
159
Hi everyone

I just wanted to share my overclocking experiences on my HP nw8000 Mobile Workstation

Made back in April 2004, this laptop has a 1.8GHz Pentium M Dothan 2MB processor, with a 400MHz FSB. The laptop's chipset is an Intel i855PM, which is only rated to do 100MHz FSB.

I tried using ClockGen and SetFSB, but they didn't support my Clock Generator (ICS950810). In fact, this Clock Generator can't be changed using software at all. So I got the datasheet and found that I could run it at 133FSB by joining pins 54 and 55.

So I did that and my CPU was now running at a 133MHz FSB, but that brought on it's own problems. First, the CPU didn't have enough voltage to run at 2.4GHz (1.34v default max), so I used a Pentium M Voltmod by shorting 2 pins on the processor to get it up to approx 1.59v. That made it fully stable.

The other problem was the RAM (DDR333) couldn't run at DDR433 which would be the new setting due to the overclock, so I had to reprogram the SPD using SPDTool to make the RAM think it was DDR266. This would trick my motherboard into running a 3:4 ratio instead of a 3:5 ratio, and that basically meant the RAM was now running at DDR354, which it can handle. I also had to reprogram the SPD to make it run at 2.5 CAS instead of 2 CAS.

So yea, everything's running good now, just thought I'd share that with you so that you know it's possible to do. I'm happy SetFSB and ClockGen didn't work, because if they did I would be happy with something like 110FSB or whatever this would reach without volt mods and with raised AGP/PCI clocks, but thankfully they didn't work with my laptop so I was forced to look for something more extreme. I think it's possible to overclock any laptop, although sometimes you'll need hardware mods (in my case everything was hardware based except reprogramming my RAM).

NW8000Screenshot.png


If you're looking for results (I know Windows Performance Assessments aren't the best benchmarks but it's the only thing I could be bothered for!):
CPU-------- 3.4 => 4.1
Memory----- 4.1 => 4.3
Graphics 1-- 3.6 => 3.6
Graphics 2-- 3.6 => 3.6
HDD-------- 4.4 => 4.4

That HDD score should increase a lot when I fit the new WD Scorpio Blue 320GB HDD tomorrow :)

I bet it's possible to mod the firmware of my ATI FireGL T2 128MB to increase the clocks etc (or I could just use ATI Tool or something) to increase my Graphics scores, but I don't really see any benefit in that for now :)
 
Nice - at least at idle. Get the eggs ready if you decide to Prime test it. Whether the eggs fry before the laptop does would be the test you need to make. :)

(I've got a 1.6 Pentium M laptop - gets plenty hot enough without any overclocking :))
 
Well actually I haven't seen the CPU go above 60'C on full load (Intel Burn Test)!

This thing is really amazingly cool and quiet, the fans don't even go up to max on Intel Burn Test... it's crazy! I think if I ran Prime 95 for around 1 hour straight then enough heat would build up inside the laptop to get the fans on full, but otherwise they just don't spin up. In fact, they rarely even turn on!

The loudest noise is the hard drive :) which I just upgraded to the WD3200BEVE (WD 320GB PATA), which seems to be the best IDE HDD on the market, so I'm hoping it'll last long! And I just bought a new laptop bag for it as well lol... So I'd be pretty disappointed if it broke down anytime soon :)
 
Congratulations on your overclock. Well done. :)

Looking at the data sheet for the ICS950810 it looks like you should be able to software select cpu clock frequencies of 100, 133, 166 & 200MHz using byte 0.

Here's a link from 2005 showing how to overclock a Pentium M Dothan by 20%. Doesn't seem to be any heat issues or news that it broke down in a short time. May be those other posters are just trying to scare you. ;)
 
actually I've had some problems with this laptop

since I put in the WD3200BEVE, my windows 7 installation kept becoming corrupt, requiring reinstall after reinstall (fully formatting each time)... I even tried XP but the same problem, XP wouldn't even install! (Disk Errors)...

either the laptop can't take it, or the Hard Drive might be faulty... I did read someone else receive a faulty WD3200BEVE, someone on newegg I think, so it could be that, because I never saw problems like this when overclocked using the Hitachi 7K60 (although I only ran for 2 days with that)...

I'm going to try this drive (the new WD) in my mum's stock Dell Precision M60 and if it gives problems then RMA it

I'm also going to try the hitachi drive again, if that develops problems then the overclock isn't good... if it runs solid then it might be the new WD drive that's faulty...

I'm in the middle of uni exams though right now so I won't be doing any of this just yet!
 
actually I've had some problems with this laptop

since I put in the WD3200BEVE, my windows 7 installation kept becoming corrupt, requiring reinstall after reinstall (fully formatting each time)... I even tried XP but the same problem, XP wouldn't even install! (Disk Errors)...

either the laptop can't take it, or the Hard Drive might be faulty... I did read someone else receive a faulty WD3200BEVE, someone on newegg I think, so it could be that, because I never saw problems like this when overclocked using the Hitachi 7K60 (although I only ran for 2 days with that)...

I'm going to try this drive (the new WD) in my mum's stock Dell Precision M60 and if it gives problems then RMA it

I'm also going to try the hitachi drive again, if that develops problems then the overclock isn't good... if it runs solid then it might be the new WD drive that's faulty...

I'm in the middle of uni exams though right now so I won't be doing any of this just yet!

This could also be caused by the Overclock if it is overclocked at the time...
 
hey everyone

I got some free time today so I sat down with my drive and tested it etc... turns out the drive is faulty

The overclock is fine, the laptop is running perfectly using my old drive... this new one seems to pass the drive tests, but it can't even manage an installation of Windows XP! (disk read errors all the time)

I've searched around and found this:
http://www.newegg.com/Product/Produ...eoOnlyMark=False&VendorMark=&Page=1&Keywords=

Read the reviews there of all the people complaining about this model arriving DoA or developing errors very quickly...

I spoke to OCUK and they were happy to replace it, but I decided I'd take this up directly with Western Digital instead... after all, it's not OCUK's fault, it's Western Digital's fault... so I don't want OCUK to have to pay the return charges to me etc etc, I'd rather WD suffer!

Anyway I'll let you all know how this all goes down!
 
I might try this with my machine as I have the 2Ghz Pentium M :p

As for talk of it dying...well it uses a two-level PCB and better is laid out than your average laptop. Oh, and the heatsink\fan are decent:

2wc2xyf.png


It is a workstation after all ;) Top of the line back in 2004. http://reviews.cnet.com/laptops/hp-compaq-nc8000-series/1707-3121_7-30583495.html

@mogli1990 - you may as well have purchased the SATA caddy and gone for a faster drive, unless the multibay battery is a big loss.
 
Last edited:
ahhh mate! if you can get that 2.0GHz stable at 2.66GHz then that would be amazing!
I'd just jump straight into feeding it 1.59volts, the heat isn't an issue!
You'd need to reprogram the SPD for your RAM though
Also you'll need to short those 2 pins on your clock generator which I mentioned to get the FSB up to 133MHz

To be honest I looked into the multi-bays but the SATA ones I found, I read that apparently they're limited to ATA33 or ATA66 or something, and they can't run at ATA133 (or ATA100) like the main drive... But I'm not sure... What have you got to say about that? If you can say otherwise then there's no reason why I won't fit a reliable 500GB SATA in here instead!
 
I've been researching and I think I'm going to get one of these:
http://www.overclockers.co.uk/showproduct.php?prodid=HD-088-SA

At 640GB, and much higher speeds (74MB/s average read as opposed to 52MB/s for the Western Digital), I think it's much more worth it... plus it's cheaper!

I'm first buying that SATA-to-PATA Caddy though to test it out using a SATA drive I have lying around... if it works full speed then I'll definitely ditch my multibay battery for this monster of a hard drive!

Besides, I think my Multibay + Primary battery get the same battery life as a brand new primary battery, so I might just get a brand new primary one... but I'll wait and see how things go with this before deciding on that!
 
ahhh mate! if you can get that 2.0GHz stable at 2.66GHz then that would be amazing!
I'd just jump straight into feeding it 1.59volts, the heat isn't an issue!

Hmm, the graphics chip is more of a bottleneck (I think there's some margin in the VRAM in particular), but it'd be interesting to see if 2.66Ghz is stable...

You'd need to reprogram the SPD for your RAM though

It's already at CAS2.5, obviously the RAM:FSB ratio will need changing.

Also you'll need to short those 2 pins on your clock generator which I mentioned to get the FSB up to 133MHz

Got a link to a guide for this? I'd rather not trash the motherboard and have to spend £50 replacing it :p

To be honest I looked into the multi-bays but the SATA ones I found, I read that apparently they're limited to ATA33 or ATA66 or something, and they can't run at ATA133 (or ATA100) like the main drive... But I'm not sure... What have you got to say about that? If you can say otherwise then there's no reason why I won't fit a reliable 500GB SATA in here instead!

Where was that?
 
I haven't yet found my GFX to be much of a bottleneck, but then again I don't really game... the most I do is Photoshop work... these old DX9 chips don't seem to have hardware acceleration in CS5 anyway...

The RAM:FSB Ratio will automatically change when you put the FSB up, provided you reprogram the SPD so that the SPD table don't display 166MHz as a supported speed (as you already know, you can check these tables in CPU-Z).

Yep there's a thread in a forum where people are shorting the pins on their IBM Thinkpad t41, uses the same clock generator. Over there they talk about using wires and pencil mods, but the pencil mods didn't work for me and the wires were way too small for me to get right... I basically "bent" the 2 legs of the IC together so they're connected now! But I didn't do this till I was sure the laptop would work at the higher FSB, cause once you do this, it's difficult to go back! Basically, to boot at 133FSB, all you need to do is switch off the laptop, short the 2 pins using a piece of metal (I used the end of a flat-head screwdriver), press the on switch... once it's past the BIOS screen, then you can let go off the screwdriver (take it away slowly so you don't short other stuff), and the system will remain at 133FSB till you reboot. Anyway, the thread (it has diagrams etc) is below:
http://forum.thinkpads.com/viewtopic.php?f=2&t=52690

Someone on the forum was saying with regards to shorting the 2 pins...
Depending on how yoru board is laid out, this may cause problems. On mine, One was pulled up and one was pulled down, so shorting them would cause both pins to be at vref/2, which would probably cause the system to not post, but would work with undoing it.. If IBM got cheap and tied them directly to vref or gnd, this would cause serious problems (ie, short between power and ground=burned trace somewhere or worse).

In short, no, this won't work.
Ignore this sissy talking about how you shouldn't short Pins fs0 and fs1 (pins 54 and 55) and how it could cause problems... it didn't cause any problems for me! I do Computer Engineering at Queen Mary (Electronic Engineering with a focus on Micro Processors) and to be honest they didn't justify what they were saying...


And about the Hard Drive, I only saw it in the following thread:
http://h30434.www3.hp.com/t5/Hardware/NW8000-NC8000-2nd-hard-disk-caddy-multibay/m-p/203847
But I also read that that's because Intel's Chip defaults the second drive on the channel to 33MHZ and how you can set it to 66MHZ within windows using a script that changes the relevant bit... however, I don't think this will be a problem for me hopefully, and I'm not using the primary drive anyway, so it shouldn't be a problem at all...
 
I was thinking more along the lines of 48-bit LBA and whether it's supported by your BIOS but seems you have found the problem is with the drive.
 
Ok, I ordered an OCZ Onyx 32GB SSD instead, and fitted it using this:
http://cgi.ebay.co.uk/ws/eBayISAPI.dll?ViewItem&item=260596971191

BUT, I didn't install it in the multibay drive (ATA Channel 1 which is limited to UDMA2 or 30MB/s in the BIOS but can be software unlocked but this is AFTER windows boots, which defeats some of the point), instead I stipped off some of the "framework" and put it in the Optical Drive bay (who needs one of those anyway!) and now it's on ATA Channel 0 running at full UDMA5 (approx 90MB/s max)!

The great thing about this, is that I get to keep my secondary battery in the multibay :D And I can also install a large IDE hard drive for storage if I ever decide I need it :)
 
Overclock 855PM on HP NC6000

Great thread :)

Just want to add that I have successfully overclocked a (cheap and old) HP nc6000 by soldering pins 54 and 55 on the ICS 950810 clock gen. So this is a notable increase in performance from from 1.6 to 2.13 @ 133MHz. I plan on getting a 1.8 pentium M off ebay and see how that goes. Hopefully from 1.8 to 2.4?

The overclock itself seemed to work, but I have had some issues with the PC333/PC2700 ram which, when flashed with SPD tool to run at 133MHz won't allow the system to boot. After trying various configurations the ram I found it will only boot with the 3:5 ratio (ram at original 166MHz), with the ram running at 226 or 452MHz. This seems to be stable although there have been ram errors when running youtube/shockwave plugins. This is contrary to whats said above, and I can't explain why? I can say though that I've been using the machine for the last 2 hours and there have been no errors, so I guess the system is stable.
 
Problem Solved!!

mogli1990(OcUK) was correct about the CAS value needing to be increased from 2 to 2.5. In SPD Tool this is not possible you can only select CAS values of 2,3,4,5 etc. so what you want to do is set the SD ram cycle time at max support to 133MHz 7.5ns AND CAS Latency (supported) to 3. This will enable the 3:4 FSB:RAM bus ratio to take effect.
 
I know I'm a little late to the game but wanted to add my experience with my wife's BenQ Joybook 5000 (manufactured late 2003):

It came with a Pentium M @ 1.4GHz (Banias, 400MHz FSB) that I naively replaced with a Pentium M 760 @ 2GHz (Dothan, 533MHz FSB).

This audience will not be surprised to hear that it only ran at 1.5GHz, after which I set out to increase the FSB speed. After tracking down some specs, I saw that it used the ICS 950810, ruling out SetFSB.

Looking at the schematic, you can see that SEL0 (pin 54) connects to VREF through a 1K resistor, and SEL1 (pin 55) to GND with a 1K resistor. This means tying FS0 to FS1 would result in VREF/2, which although might have worked I didn't want to try. Instead, I decided to close resistor R262 using a conductive pen.

In tests I did on paper, plastic, and scrap PCB, the resistance of the conductive ink didn't drop until it had been drying for an hour or more under a 40W bulb (which I guess is why they also sell a fast-drying version).

I was a little freaked out that the BIOS still said it was a 1.5GHz processor, but once Windows starts it reports (both Computer -> Properties and using CPU-Z) that it was now running at 2GHz.

I ran Prime95 and it immediately failed with memory errors. I used mogli1990(OcUK)'s recommendation in SPDTool (Max Supported = 7.5ns / 133MHz and CAS Latency = 3) and now everything's working great.
 
Last edited:
Back
Top Bottom