On Sat, 22 Jun 1996 14:01:42 -0700 
 "Michael L. VanLoon -- HeadCandy.com" <michaelv@HeadCandy.com> wrote:

 > Do you have a decent heat sink and fan on the chip?  Does the heat
 > sink make _uniform_ contact with the chip across the entire surface?
 > If it's a heat sink without adhesive, go to Radio Shack and pick up a
 > little tube of their silicone heat sink compound (little blue and
 > white tube) and carefully spread that across the entire surface of the
 > chip (just a thin layer) and carefully but firmly reapply the heat
 > sink.  It's kind of messy, but it's worth it.

I'd guess it's not a heat problem.  When I filed my PR in April, I 
noticed the problem after powering the machine on after 8 hours of downtime.

 > Assuming the problem isn't the CPU, you might also try a memory bus
 > speed of 60MHz, instead of the ~66MHz you're using with both 166MHz
 > (~66.7 * 2.5) and 133MHz (~66.7 * 2).  This is in case your RAM just
 > can't keep up.  Try running at 150MHz (60 * 2.5) or 120MHz (60 * 2).

I assume there's some way to change this via the BIOS or a jumper on the 
mainboard?  Need to install a new crystal?

 > On my ASUS P55TP4N, the motherboard L2 cache is write-through, and the
 > Pentium's on-chip L1 cache is write-back.  I would expect this to be
 > the typical arrangement since the Pentium is designed with an internal
 > L1 write-back cache, and having two write-back caches makes the
 > coherency logic much more complicated.

Ah, ok.  So, all of it should be dealt with in hardware...software 
shouldn't have to touch it?

 -- save the ancient forests - http://www.bayarea.net/~thorpej/forest/ -- 
Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939