PowerMac 6100 Graphics Performance
First, I ran Ziff-Davis' standard MacBench utility, version 3.0. I ran the standard Graphics Mix test with varying configurations at 8 bits and 16 bits of color depth. The base system is profiled in the legalese that follows, and is almost identical to the 6100/60 that Ziff-Davis uses as the 100% benchmark system at 8 bits. My measurements are slightly faster, which I believe results from having Virtual Memory turned off, and having a larger disk cache (256K vs. 96K).
Options added in subsequent tests were a 256K Level 2 cache from PowerLogic, the HPV card from Apple, and a 78 MHz Rocket Socket clock accelerator from Output Enablers.
Using the same configurations, I played the most excellent game, Marathon 2, and recorded the frame rates I observed.
While I also play DOOM, Hexen, and Descent on my machine, Marathon 2 is the only game I've got that will calculate frame rates for me. To get frame rates in Marathon 2, press shift-? while you're playing. The frame rate will be displayed in the lower left-hand corner. See the hi-res screenshot for an example. Here are the results:
| fps Low-Res 256 Colors | fps Hi-Res Thousands |
|
|---|---|---|
| Base System | 15.38/32.43 | 4.17/ 8.45 |
| cache | 16.22/33.33 | 4.84/11.65 |
| HPV card | 16.35/32.43 | 6.59/14.59 |
| cache and HPV card | 20.69/33.33 | 7.69/15.79 |
| cache, HPV and clock chip | 25.69/33.33 | 8.48/19.67 |
To see an screenshot of Marathon 2 at Low Resolution, 256 Colors, click here, and to see a screenshot at High Resolution, Thousands of Colors, click here.
Note that these are the extremes of the frame rates achieved. Maximum rates were achieved when there were relatively few textures and no live bad guys in view. The minimum frame rates were observed during firefights with lots of things happening at once, and the frame rates didn't stay down that low for very long.
The average sustained frame rate was usually about halfway between the slowest and fastest, or maybe slightly faster. So in Low-Res, 256 color mode, I usually got around 25 fps or better.
Unfortunately, what the numbers can't convey is the qualitative feel of how the game plays with the various configurations. It's my personal preference, but I always play with the largest viewable area I can. I consider a sustained 25 fps to be pretty darn good, and a sustained 20 fps is acceptable. Below that, the motion is jerky enough to be distracting, and important things (like getting shot) can happen in the interstices between frames. Anywhere south of 15 fps is ridiculous, and looks like a time-lapse movie.
Unfortunately (again), what the data shows and what my anecdotal experience bears out, is that Marathon 2 doesn't run very well on my machine at its maximum settings. Sigh. All those fantastic graphics, and I'm missing out. Then again, from the test results, I'm guessing that it would take a fairly fast 604 to run well at the highest resolution.
On the plus side, the data also show that Marathon 2 runs great on my machine at a lower graphics setting. Even at Low-Res, 256 colors, Marathon 2 blows away Doom for visual impact. The graphic resolution is far less important than the frame rate and and the action. Furthermore, a stock 6100/60 can achieve an impressive 30+ fps when you're cruising around. In Marathon, this happens quite a bit as it's more puzzle-solving and exploring oriented than the shoot-em-up Doom. Kudos to Bungie for writing an engine that cranks so well on an entry level machine!
Back to the PowerMac 6100 Graphics Page
Back to the PowerMac 6100 Upgrade Page
This page is maintained by Steven Kan. Feedback? Corrections? Mail me @ steven@kan.org
Revision 2.0
Thanks, and please come back soon!
