Thinkpads Forum

I've been using my upgraded TP 600E for about half a year now, and I'm pretty satisfied with it. There are two things that annoy me.
One of the things is that, at rare occassions, the graphic chip goes ballistic and glitches everything on my screen. That error persists even after rebooting, but disappears when I switch drivers. So it's not really a big problem since it can be fixed. In fact, as an artist, I find the glitches appealing, and would like to know a way to summon them. I'll also upload som screenshots to my log soon, I'll post a link when I'm done.

The CPU randomly starts at either 500*1.08 MHz or 650*1.08 MHz, and has L2 cache disabled. This is of course easily cured with deepsleep and powerleap. But I don't like the fact that pleap shows a window when booting and that DS requires you to answer push ok to a dialog box. I could modify deepsleep, but if anyone knows how to activate the PIII's L2 cache using asm code, I'd be happy to know.
I'd also like to activate L2 cache as early as possible. It's probably too much to hope for that there'll be a BIOS hack, but at least I'd like to do it already at boot time. (Which I may be able to do a hack for if I just find out how to activate the L2 cache)

The CPU gets pretty hot, 90° C at absolute peaks, typically 60-75° C at average load, according to mobilemeter. Is this acceptable? so far it has kept on working, but will the heat do any long-term damage on the CPU? Or was it designed to tolerate such heat?

The 32 MB onboard RAM seems to be working stable at 100 MHz.

And so a question about what the subject line says. About 50% of the time I get an error 108 when booting. According to the documentation, error 108 means "Timer bus test failure", which sounds reasonable. However, I don't think anything is really different when the test fails, just that test is executed in the wrong time, so to say.
So, is there a way to hack CMOS to make Easy-Setup ignore that test?

high quality pc133/100/66 low density ram on them 600e
remmember bx chipsets are capable of 150fsb+and outperform all
the newer pc100,133,810,815,845 rdram600,700,800 chipsets!!
even themsome pentium 4 machines with sdram,ddrr1600 and 2100running dud batterys (lowx speed modes).
so i doubt its a hardware issue but rather a bios issue
governed by corporate body profit manipulation incentives,
like intels rdram hoax(more where that came from)but thats another story.
one chap got his tp600e over 139fsb with modded clock generator
i found onboard ram on 600 has more aggressive timings than the 600x + it certainly out benched the 600x too on identical host speeds.

you may try and deviate the memory sticks
+or disable the onboard ram to lessen the 108 host bus alert
personally i found using a 750+up mmc 2 cpu somehow vritually eliminates the effect,i noticed that if i remove a battery or swap a drive in the multibay that makes it appear more than not
the ultimate cure has not been established yet but these are how
i saw the events in regardas to 108 boot prompt.
if nothing the above helps than you can always revert to 105 fsb
and see if that helps.
*now about the l2 cache ,assumming your running windows
theres a neat program called cpu msr v090 (google it)
that little program can enable the cache and other parameters
in windows ,however you have to save the modded file to my document,and drag copy it to c/documents and settings/allusers/start menu/programs/startup*folder.
for it to autoload in windows xp.
it works so nice you see no prompts ,no delays infact can't tell
the diff from 600x .conversely it took me about half an hour to sort out all the pros and cons including uninstalling plleap prior to hand.