Speed Tests: Apple Users Bite Back
Few things rile Mac users more than speed comparisons with Windows machines. At least that’s what my mail suggests. Scores of responses to my last column took me to task for my conclusion that the fastest desktop Macs and Windows PCs track pretty closely in performance. (In case you missed it, I also concluded that the PowerBook 3400 is the fastest notebook computer.)
Most of the correspondents relied on performance testing by various magazines that, unfortunately, don’t tell the full story. (For the record, until this year I managed Macworld’s features section, where I spent several years overseeing the testing of Macs.) At the risk of extending what may be an exercise in testosterone-induced chest-beating (nearly all of those who disputed my conclusions were men), this column offers a rejoinder:
Taking the Vendors’ Bait: In its July issue, Popular Science anoints the Power Macintosh 6500/300 as the fastest home computer. While the 6500/300 is a great box, the article shows why it pays to ignore computer tests from magazines that don’t specialize in computing. (Consumer Reports has also done some truly awful computer comparisons.)
The article adopts computer vendors’ designations for “home machines” as a basis for comparison. These designations reflect strategies for how to hook buyers, more than intrinsic home qualities. Check out the computer magazines alongside Popular Science on the newsstand. Faster and less expensive PCs--and Macs--abound.
Selective Testing: Many readers pointed to testing by MacUser and Byte magazines as evidence of the Mac’s overall superiority. Each magazine tested a single product, Adobe’s Photoshop image-editing software. In MacUser, a 200 MHz Mac beat the top-flight PC based on Intel’s flagship 266 MHz Pentium II in 10 of 17 tests.
Seven-month-old testing by Byte showed the Mac as the clear winner in two categories, a slight winner in one, a slight loser in two. In Byte’s June issue--strangely ignored by those who wrote to me--a top Pentium II and a top Mac ran “neck and neck” on Photoshop tests.
These results hardly show unequivocal superiority for the Mac in graphics performance, let alone as a general-purpose computer. (The misleading cover type for MacUser’s August issue--”Pentium II: Still No Threat to the Mac”--doesn’t help either.)
*
Wrong Face-Off: Others cited the July issue of Macworld, which designates the Mac as the fastest desktop computer after it beat a PC in three categories, tied in one and lost in two.
Credit Macworld with comprehensive testing. But it failed to test the appropriate PC. It compared a 250 MHz Mac against a PC based on a 233 MHz Pentium II. Yet PC World published results for a Pentium II-266 about six weeks before the Macworld story appeared, and most major PC magazines published 266 MHz results when the Macworld article appeared.
An average Pentium II-266 system runs about 5% faster than a 233 MHz unit using business productivity software, and about 11% faster for multimedia--close to the overall difference between the top Mac and PC in Macworld’s tests. If you’re trying to pick the fastest personal computer on the planet, such details matter.
(It’s worth noting that Macworld correctly points out that the fastest Macs still cost substantially more than the fastest PCs. And that painful gap is growing again with the release of Windows PCs based on new chips from Cyrix and AMD that shave hundreds of dollars off the price of high-performance computing.)
Other readers reached back a year or more for data to rebut my premise. In many cases, the figures show Macs using PowerPC chips running faster than Windows boxes. But overall, PowerPC- and Pentium-based machines have been competitive over time, with the Mac keeping a healthy edge in some areas, trailing in others.
But so what? For most users, splitting hairs on speed is a waste of . . . time. Speed is only one aspect of a person’s computing experience, and rarely the most important. Even for professional publishers, artists and others who need the fastest machines available, features, quality and usability matter at least as much. Those people don’t need a sales job on how fast the Mac runs a Guassian Blur in Photoshop. Like the design department at PC World, they already use Macs.
*
Charles Piller can be reached via e-mail at cpiller@aol.com
