The iPad’s Performance: Weak, or Strong for its Size?

An interesting development in the world of testing geekdom: Primate Labs has delivered a version of its cross-platform benchmarking software to the iPhone platform, and what a door of wonder that has opened up.

We’ve always liked Geekbench because it does as good as job as you’ll see in comparing computing devices on an apples-to-apples basis regardless of whether they run Linux, Windows XP, Vista or 7, or Mac OS X. The trouble for iPhone fans is that, until now, there has been no benchmarking software that could provide us with a relevant comparison between the iPad and legacy computing devices.

Well, folks, for the low, low price of $1.99 from the Apple iTunes App Store, you can now download Geekbench 2.1.9 for iPhone OS and see for yourself how Apple’s wonder products stack up head-to-head with PCs. (At least on a system performance basis.)

The app measures integer performance, floating point, memory and stream.

We downloaded Primate Labs’ app, and ran it on both an iPad and iPhone 4. The results:

The 32GB iPad, running iPhone 3.2, chalked up a score of 436. The iPhone 4, running iPhone 4.0.1, scored a 362.

For the iPad, that puts it at about 25 percent of the performance for a similarly priced notebook PC, and about 20-25 percent for the iPhone.

So on a simple, hardware-focused performance measurement, that’s not great. But like most benchmarking techniques, the numbers only tell part of the story. For example, iPad gets 12 hours of battery life out of the box. But it can’t multitask. But it’s feather-light and a breeze to carry around and actually use.

The 436 Geekbench score means that the iPad provides about one-tenth the compute power (as measured by Primate Labs’ software) of today’s notebook PCs and even less performance than a lot of netbooks. But Apple crams an awful lot of “wow” into that 436 score.

The bottom line: Apple hasn’t designed iPhones or iPads to be performance devices and that was clear even without benchmarking software. Geekbench just confirms that. But those devices perform in “intangible” ways - - ways you can’t benchmark with a mathematical equation. Fortunately, there is another time-tested benchmark for how strong the iPad performs: sales. Apple had sold 3.27 million iPads through the end of June.

So there are benchmarks, it seems, and there are benchmarks.