Hardware Freaks with too little time on their hands

• blah
  • blah

In order to ensure that your test results are accurate, you'll need to download the benchmark file, which you'll find located right here. The reason you need to use this work unit is due to the fact that no two work units are the same. Therefore, each work unit will have a different completion time but by using one file, we can ensure that we get accurate, reproducible results.

In order to use this file, make a copy of your current SETI directory. Unzip the work unit, replace the work unit file in the copied directory with the unzipped one and delete the result header and the state file. The reason that you have to delete the result header and the state file is to ensure that your computer starts as the beginning of the work unit, instead of in the middle. Also, make sure that you append a -stop_after_process onto the CLI (Command line client), or change your GUI's connection option to manual connection. Otherwise, your computer will connect to the server and download a new unit, deleting your CPU time. 

When running the benchmark it's recommended that you not do anything else on your computer, since this will affect the times. Therefore, if you plan to do some web browsing or a little Q3/UT loving, I suggest you stop SETI and restart after the violence is over. Also, make sure that you disable the System Maintenance  Manager (Task Scheduler) on Win9X and Find Fast. Defragmenting while you're processing the unit is a definite speed hit!

Problems with the submissions page?
You can email your details:
CPU
clock MHZ
FSB and actual memory speed (new: see submission page for details).
RAM
OS
CLI version,
number of clients running
cpu_time=(line 18 of the result.sah file)
name you wish to be known by (email addresses kept confidential)
comments.
Please attach 'result.sah' (5kb) from the completed unit (this is mandatory for inclusion in the results table) and send to Roelof.

Last updated 13 February 2001.