Can a PM45 run stable at 1600 MHz FSB?

Hi everyone,

My 2xElitebook 8530's has been running awesomely for nearly 2 years since I got it, but they're pretty old and I want to squeeze as much performance out of them as possible, since I'd like to delay getting a new laptop and get a good monitor right now.

So I'm planning on PLL modding them. But I need some advice - how will the PM45 northbridge behave at 1600 MHz FSB? Will it be stable/hotter? If anyone has some hands-on experience, please chime in!

Most people PLL mod from 266 to 333 MHz FSB, which seems to work without issues for years on the PM45, but if 1600 MHz is also stable and not much hotter, I'd go for it. So far I've seen only one example of 1600 MHz FSB overclock here: http://forum.notebookreview.com/gat...ly-overclocked-p7805u-cpu-22.html#post6707815

I've got cooling, soldering, RAM speed and processor multiplier (hopefully) figured out - I'm just deciding whether I go for 1333 MHz or 1600 MHz FSB.

My processors are a QX9200 and a T9900.

 

The problem as always when running a chip above its rated specifications is stability. Heat may not be your problem. When overclocking desktops my long-term stability was dependent on getting the voltages correct (especially CPU); and notebooks of course don't permit voltage adjustments. When overclocking desktops I also disabled all the power saving features which would otherwise jeopardize stability; again notebooks don't always allow those kind of adjustments.

Also consider RAM - what frequencies does the notebook support? Keep the FSB: DRAM ratio something reasonable for maximum stability i.e. 1:1 is ideal but 3:2, 5:4 works, &c.

The good thing is you have two high-end CPU's; Intel cherry-picks those because they're stable at higher frequencies.
It's a risk no matter how you view it. Personally I think 1600MHz FSB sounds too extreme (then again I had chips running at 150% of their rated speed Prime95 48 hrs stable).

 

Thanks for the reply, Charles!

It does look like 1600 is too much for the north bridge itself (probably would work if I could raise voltage ). RAM is not an issue, since I can flash it to lower clocks/timings (i've got DDR2-800 - flashed to 667 it would run at 830Mhz @ 1333 MHz FSB mod, 3:2 ratio I believe). I've read other people's reports on 1333Mhz FSB mods on the PM45 chipset and it seems that their laptops run without issues long term at that frequency.

 

You're welcome.
I'd keep the RAM timings at something relaxed/"mainstream" as that could contribute towards stability. I experimented with a Core 2 Quad using various memory timings; going from CAS5 to CAS4 made a next-to-nothing difference (less than 1/10th of a percent). I haven't bothered with memory timings since then.

Just make sure you're Prime95 stable. I also used memtest86+ for memory testing if I changed memory timing or voltage. Minimum 24 hours error-free for both (IMO). And of course keep an eye on the temps.