The OCUS Benchmark version 3.3
(for Releases 2000i, 2000i2, 2001)

Wildfire support has been removed.
The results of Wildfire were incorrect and much too slow.
For more info check the comparison page

In Short
Requirements
History
Download
README
R2000i compared to newer releases
R E S U L T S
- Last Modified: -
The OCUS Benchmark is a do it yourself benchmark.
Version 3.3 consists of:
  • a part,
  • an assembly
  • 8 trailfiles,
  • configuration files: config.pro, ocus.scl, ocus_tree.cfg
  • the README file with the instructions.
  • For unix users the script ocus_bench to count the results has been added.
  • For windows users a count utility has been added.

In short:

You download the benchmark (read the README file) and run the trail file on your own computer. Count the results with the count utility. Send the results to me by sending the time stamps to my web-based count tool.

You can view the results here.

What does the trail file do?

It opens the generic part and assembly (consisting of many instances of the part) and does many repeated actions on both and on a created drawing of the assembly. In between each repetition of tasks a time stamp is created.

Which tests are performed?

Requirements:

The normal (large) version needs:
  • RAM minimum 512 MB
  • Available disk space: 180 Mb
  • Pro/E: R2000i (preferably=fastest) or R2000i2, R2001 (need 576Mb RAM)
  • Runtime: Anywhere between 20 minutes and 2 hours
  • Trail file: ocus_bench_31.txt
I prefer you run the normal (large) version, but to facilitate users who don't have a machine with 512 Mb RAM you can also download a light version of this benchmark with a smaller assembly but larger trail file:
  • RAM minimum 256 MB
  • Available disk space: 30 Mb
  • Pro/E: R2000i (preferably=fastest) or R2000i2 or R2001
  • Runtime: Anywhere between 10 minutes and 1 hour
  • Trail file: ocus_bench_31_light.txt
If your machine has enough RAM to run the large benchmark then I would appreciate it if you also run the small version and send in those results as well.

The preferred release to run the benchmark is R2000i, but I also included R2000i2, R2001 and Wildfrie trailfiles. Note however that R2000i is a lot faster than the other releases. So be carefull comparing R2000i with R2000i2 or R2001 results.

History:

  • 20030822 3.3a Updated README file for the new count tool.
  • 20030605 3.3 Copy of the 3.1b version with the new count tool
    Version 3.2 with Wildifre trailfiles has been withdrawn. Wildfire results were incorrect
  • 20030521 3.2 Added Wildfire trailfiles. Added new windows count utiltity.Withdrawn
  • 20011121 3.1b No changes: Added R2001 trailfiles.
  • 20010327 3.1a Minor changes. Fixed ocus_c31.bas file (DISK total was not correct with NT count utility). Added FLAT to IGES_OUT_ASSEMBLY_DEFAULT_MODE config.pro option.
  • 20010108 3.1 Preferred Release to run the benchmark is now R2000i (fastest results).

    • Seperate R2000i2 trailfiles are now available. The prvious version did not run correctly on R2000i2. (If you've allready downloaded the previous version (3.0a or 3.0) you don't need to download this version if you're only running on R2000i.)
  • 20001215 3.0a ocus_b30a Fixed count_top.htm file.
  • 20001201 3.0 It is now a lot more demanding: 
    • the previous version was not very good in testing the graphics card, 
    • the fastest machines we're so fast that the results became inaccurate. 
    • A lot more tests have been included (like advanced shading and drawing mode) 
    • There is now also a light version, for computers who can't handle the normal one. 

Download:

I encourage you to perform this benchmark on your own computer. Check if it meets the above requirements and download the benchmark here (large and small version included):
v3.2 has been withdrawn.
The results of Wildfire were incorrect and much too slow.
For more info check the comparison page
  • I prefer you send your results to me with my web-based count utility

    • This will be automatically started when you count your results with the count program of the benchmark.

Also carefully read the README file.

Other benchmarks on the net: