As is standard for my recent reviews, I ran timed applications tests in Bryce 2 and Infini-D 4.0. I used standard sample files provided with the applications. I chose these two applications because:

1. There is no human timing errors possible since the application reports the total time to complete the test.
2. They represent CPU intensive tasks of varying durations.
3. They are very popular and therefore are likely to be used by many people reading this review.

I'm sure some of you will ask why I didn't run the full range of tests (as done with many other G3 cards) but to be honest I'm buried in work and email. The two tests I've run were very CPU intensive and of sufficient duration to be valid comparisons of performance. In general they also are less influenced by other system variables such as disk speed (unlike Photoshop which hits the swap disk on almost every operation, even with large amounts of RAM allocated to the program).

I know many of you would like to see more but I'm doing the best I can at the moment considering I'm answering about 6 hours of support mail a day, 7 days a week.

I used the same Infini-D tutorial file (chapter 7 completed) rendering test as I've done in past reviews. Rendering quality was set to Ray Trace, medium anti-aliasing, shadows on, patch detail low.

The table below shows the results at both 266/266 and 300/300 speeds. The PowerForce 220/110 and the XLR8 266 G3 cards results are also shown for comparison. Other G3 CPU cards I've reviewed were tested with Infini-D 4.0 and showed a large difference in times from v4.01, so I felt the version change made any comparisons to those results invalid.

Although some may say it's not "fair" to compare the $499 PowerForce G3 220/110/512K card to ones that cost more than twice the price, I feel it is a valuable piece of information for those considering cards at both ends of the price spectrum. It also graphically illustrates the differences that larger/faster caches and CPU speeds produce in real applications. You can decide for yourself it the difference in performance is worth the cost. For more details on the other cards shown, read my XLR8 266/177 and PowerForce 220/110 reviews.

I ran tests in Bryce 2 using the sample scenes in the KAI folder called "Magical Knight I" and "Alexandria II". Bryce 2 was left at the default memory allocation of 16980k.

Since my version of Bryce 2 is the same now as it was for all the other CPU card tests I've done, I listed more cards here for comparison, including the fastest 604E I've tested, a PowerLogix PowerBoost Pro 233 running at 263 Mhz with a 1MB L2 Rapidcache.

(For reviews of all the cards listed - see my CPU card review list at the main site.)

Since applications performance was very good, I rated this category a 9.

The next page describes the software cache control supplied with the card. Or you may use the links below to jump to a specific page.