Hmmm... So are you saying that "GROUP BY" operations use the CPU cache to a larger
extent than other operations? (Is this documented somewhere?)
Don't forget that the other 80-odd operations listed in the benchmark are actually
faster on the new machine. It's odd that only these "GROUP BY" type
Kevin Cowley wrote:
Don't tell me - you upgraded a PIII server to a PIV server?
The cache on the PIV is les than half the size of a PIII.
We've hit this problem with our own apps that by the way they operate cache
a lot of data. A PIII 1.4GHz will match a 2.4 GHz PIV.
Yes, MySQL code is written wit
Don't tell me - you upgraded a PIII server to a PIV server?
The cache on the PIV is les than half the size of a PIII.
We've hit this problem with our own apps that by the way they operate cache
a lot of data. A PIII 1.4GHz will match a 2.4 GHz PIV.
Kevin Cowley
R&D
Tel: 0118 902 9099 (direct li
