MPQS Checkpointing?

[Andrew Brown <abrown@genuity.net>]

In the near future, I will be doing some semi-large (80 to 100 digit) 
factorizations using PARI's excellent implementation of the MPQS.  I am
wondering if there is any way to checkpoint the process to disk, as these
factorizations can run for several weeks.

Also, using the PARI library, is it possible to split up what ranges or
polynomials are used during the sieving process?  Can the elimination
phase take advantage of multiple CPUs?  It would he helpful if there were
ways to distribute the workload. 

Thanks in advance.

- Andy Brown