12/29/06

Benchmarking!

I know this is a bad idea. The IIcx is no speed demon - why do I need to quantify it with actual numbers? Well, there's no good answer to that. We're gonna do it anyway.

I'm using Speedometer 4.0 for the benchmarks. I would have used 3.x, but I couldn't find it online anywhere. Anyway, I paired my IIcx against the other machines I had available to me - a IIci, a IIsi, and a Quadra 950. Of course, the Quadra's numbers don't really tell you anything beyond showing the ridiculous dominance the '040 has over the '030. All benchmarks were executed under a bare bones System 7.1. The IIci had the Apple cache card installed, and the IIsi had the Nubus adapter, which provided an FPU. All graphics benchmarks were done using internal video, except of course the IIcx, which used the Toby card. They were all tested using an 8-bit pixel depth. Like all Speedometer 4.0 scores, a Quadra 605 is the baseline - 1.0. Here are the results:

It's no surprise that the IIcx got its butt kicked. But the graphics performance isn't really any worse than the IIsi, although the IIci was almost twice as fast. The CPU and FPU scores were pretty much as expected - the scores increased linearly with the clock rate of the processor (the IIcx, IIsi, and IIci have a 16Mhz, 20MHz, and 25MHz processor, respectively). The IIci did just a bit better because of the cache card. Naturally, the '040 showed a pretty crazy leap in performance. I didn't indicate disk scores because I used the same drive in all the '030 machines to keep the software consistent - an Apple branded IBM 160MB drive. The disk performance was the same across the board - about .661 using a 128KB disk cache.

So what does this prove? Absolutely nothing, but I was curious nonetheless. LowEndMac has a pretty comprehensive listing of Speedometer 4 scores if you're interested. As always, these scores aren't necessarily a realistic account of performance, but it's good enough for my purposes.

Back to the main page