G
Geoff Briedé
Hello All,
I'm a corporate IT git and have inherited a bunch of dual Xeon workstations
to look after - various IBM M and Z Pros and Dell Precisions - all with
either 400FSB or 533FSB and SCSI hard disk arrays that are either LVD 160 or
LVD 320.... nice comps...... all of them.
In order to answer some rather pointed questions I need to get a REALLY good
benchmarking programme to run tests on these boxes.
I've been searching the net and trying various programmes and SiSoft Sandra
seems to be the best so far, however it incorrectly reports mainboard data,
telling me things like the mainboard and one of the processors are both
running too hot (they are NOT - I've had my hand on both heat sinks and
can't tell the difference, even though they are supposed to have an 80
degree (!!!) difference.
I'm worried that if this programme can be so WRONG about so many things, can
it be trusted to report accurately on the other data; the data I really need
to be accurate? No, of course I can't.
So, I appeal to the pundits at this forum to recommend benchmark-type
programmes in the hope that I don't have to go through another ten until I
find one that seems good, but isn't.
Recommendations, please?
TIA,
- Geoff
I'm a corporate IT git and have inherited a bunch of dual Xeon workstations
to look after - various IBM M and Z Pros and Dell Precisions - all with
either 400FSB or 533FSB and SCSI hard disk arrays that are either LVD 160 or
LVD 320.... nice comps...... all of them.
In order to answer some rather pointed questions I need to get a REALLY good
benchmarking programme to run tests on these boxes.
I've been searching the net and trying various programmes and SiSoft Sandra
seems to be the best so far, however it incorrectly reports mainboard data,
telling me things like the mainboard and one of the processors are both
running too hot (they are NOT - I've had my hand on both heat sinks and
can't tell the difference, even though they are supposed to have an 80
degree (!!!) difference.
I'm worried that if this programme can be so WRONG about so many things, can
it be trusted to report accurately on the other data; the data I really need
to be accurate? No, of course I can't.
So, I appeal to the pundits at this forum to recommend benchmark-type
programmes in the hope that I don't have to go through another ten until I
find one that seems good, but isn't.
Recommendations, please?
TIA,
- Geoff