Extremely high context switching

I have just moved a database to a new server and new version of Postgres
(7.1.2 to 7.3.4). The new server is much more powerful (2GB RAM, 2xXeon,
RAID) than the old one (1GB RAM, 1 CPU, single drive) and I would expect at
least twice the performance if not more, however, it is constantly showing
high load and very high context switching, even outside peak hours. Here's a
portion of vmstat output:

procs                      memory      swap          io     system
cpu
 r  b   swpd   free   buff  cache   si   so    bi    bo   in    cs us sy id
wa
 3  0      0 128000  68040 1680488    0    0     5    23   29    17  1  0 23
1
 3  0      0 128028  68040 1680488    0    0     0     0  114 242523 43 16
41  0
 3  0      0 128112  68040 1680488    0    0     0     0  114 239304 42 16
41  0
 4  0      0 127556  68040 1680488    0    0     0     0  166 211554 43 27
30  0
 2  0      0 127476  68040 1680488    0    0     0   448  220 194517 50 14
35  1
 3  0      0 127200  68040 1680488    0    0     0     0  142 190350 50 14
36  0
 3  0      0 127864  68040 1680656    0    0     0     0  182 209546 55 19
26  0
 3  0      0 127424  68040 1680752    0    0     0     0  162 155351 52 17
31  0
 3  0      0 127856  68040 1680848    0    0    32     0  137 198027 59 18
23  0
 2  0      0 127588  68040 1680920    0    0     0   548  238 220461 38 16
46  0
 2  0      0 127864  68040 1680920    0    0     0     0  123 193224 27 12
61  0
 1  0      0 127960  68040 1680920    0    0     0     0  135 220137 38 16
47  0
 5  0      0 126472  68040 1680920    0    0     0     0  152 131971 34  9
57  0
 4  0      0 127016  68040 1680920    0    0     0     0  151 45508 32  3 65
0
 3  0      0 126872  68040 1680920    0    0     0   120  138 47118 31  5 64
1
 1  0      0 126988  68040 1680920    0    0     0     0  142 41675 30  3 66
0
 1  0      0 126860  68040 1680920    0    0     0     0  132 64751 29  4 67
0
 1  0      0 126896  68040 1680920    0    0     0     0  164 43179 23  3 73
0
 2  0      0 126872  68040 1680920    0    0     0     0  115 42461 24  3 74
0
 1  0      0 127452  68040 1680920    0    0     0   416  153 40434 24  2 73
1
 1  0      0 127436  68040 1680920    0    0     0     0  123 42127 23  4 73
0
 1  1      0 127372  68040 1680960    0    0     8     0  146 74913 28  5 67
0
 1  0      0 127004  68040 1681072    0    0     8     0  148 47233 26  7 67
0
 1  0      0 126988  68040 1681072    0    0     0     0  123 47676 30  6 64
0

When the system is idle, cs is ok (less than 50).
Is this postgres problem or a system problem? Unfortunatelly, too many
parameters have changed, but I'm hoping for any sort of suggestion as to how
I can fix this.

Thanks,

Bojan