So I'm in New York to Watch the Video Downlink for the Rollout of Apple's Newest Powerpc 604-based Computers as well as a new color laser printer. As the show progresses, I start to think that Apple has the right idea here: Show equipment for the graphics and publishing community that easily outperforms anything else available. Everything is going smoothly until I see the big blunder. Just when Apple can deliver a coup de grace to the PC, it drops the ball.
During the presentation (and, by the way, let's get some personality transplants for these "new" Apple spokespeople!), Apple decides to put a new 604-based Power Mac up against a dual-processor Pentium system for some heavy-duty Photoshop special effects. It's not known if the concentric-multiprocessing Pentium system can actually optimize the software -- probably not. Whatever the case, the Apple folks begin a Gaussian blur or some equally arduous process on both machines at the same time. Tick tock tick tock. Everyone waits in anticipation. Tick tock tick . . . bingo! The Mac finishes first! Yes! Everyone is ecstatic. Now when will the Pentium finish?
Holy Toledo, they cut away from the Pentium and start spouting some gibberish about the new machines. Hang on, I'm thinking. Does the Pentium take all day? Does it finish one second after the Mac? What? We are left in the lurch by this boneheaded change in direction. Doesn't anyone at Apple have a sense of the dramatic? Don't they think someone might want to see exactly how slow the Pentium really is? I suppose the Pentium's time to the finish line will remain a secret forever.
This incident really upset me. Especially since the real data coming out regarding Photoshop performance indicates that a 604 can run math-intensive functions two times as fast as a Pentium can, owing to an incredibly superior floating-point system within the chip. I'm further distressed by publications such as Microprocessor Report (June 19, '95, page 12) that ignore these realities and look only at raw integer performance in chip comparisons of the Pentium and the 604. That particular story has no comparison of the seemingly more important floating-point performance of the two chips. I say "seemingly more important," because people are using these fast systems for graphics, not just for word processing and data management. PC WEEK tests indicated that the 604-based Power Macs are twice as fast as Pentium-based PCs for graphics apps. This finding was further confirmed by PC Magazine. If someone can reduce a ray-trace rendering time from four hours to two hours, don't you think that's an advantage worth promoting?
Of course it is, but Apple doesn't seem to understand this at all. And Apple's lack of understanding was epitomized by the failure to ridicule the Pentium as it labored over that Gaussian blur onstage.
Advice to Apple: Advertise performance. Put the Power Mac side by side with a variety of typical PCs. Show the performance differences on TV and in print media. In many countries, this kind of comparative ad is illegal, because it shows distinct advantages. Apple should capitalize on its clear performance advantage. Get a clue!
In the past, and up until Windows 95, there was still a leftover perception (not totally exploited by Apple, either) that the Mac was easier to use and set up than a PC. The gap insofar as ease of use is concerned has narrowed. The Mac is still easier to use, but not by as much as some people would like to believe. Worse, the Mac's versatility is less than that of a PC, because most PCs have some form of multitasking. Combine that with cheap peripherals, and what can Apple offer in response?
Performance. Performance has always led the desktop-computing parade. That's why there's such an overall panic to increase performance in every way, all the time. Processors are faster. Hard drives are faster and bigger. Everything improves at breakneck speed. When Apple was using the 68000 chip family and the PC world was using the 80X86 family, performance was very close. The 68000 had some minor advantages in the early going, but nothing spectacular. But no matter, Motorola seemed to get distracted and Intel suddenly shot ahead with the 486 and the Pentium. It was obvious a few years ago that Motorola was falling behind Intel in the performance game. Then along came the PowerPC chip. It's now clear that the PowerPC is the superior chip in integer and floating-point benchmarks -- floating-point tests in particular. And there's no reason not to believe that this chip family will hold steady and maintain the edge.
Given all that, how do Apple's leaders plan to exploit this clear advantage? If they have a plan, it's a well-kept secret. Kind of like the secret length of time it took that dual-Pentium system to finish the calculation during the latest Power Mac rollout.
