Wow - I'd love 8:1 performance improvement, across the board, with nothing
more than the s/w upgrade - all managers who repeat such sweeping statements
(that they heard somewhere) should undertake a vigorous online study course
at www.dilbert.com.

I can think of a case where this *might* be true and that is using the new
RESOURCEUTILISATION parameter on the client. If the client has 8
filesystems, oodles of spare CPU and a huge network pipe to a previously
understressed server then it could happen..... Of course we all did this
before but now at 3.7 we can schedule it from the server without multiple
licensed nodes etc.

I don't think we've got adaptive differencing yet - in a future PTF was my
understanding. This plus the ability to cut CDs on the server (except for
mainframe based servers) means we could have a pretty sensible laptop backup
strategy.

Restore performance improvements = none apparent.

Regards
Alan