Single page Print

Performance
We've cobbled together a new suite of tests to gauge performance with the sort of basic everyday tasks that occupy most ultraportables. Since mobile browsing is the raison d'être for most of these systems, we'll kick things off with a couple of browser tests running in Firefox. The first is FutureMark's Peacekeeper benchmark, which the company says tests JavaScript functions commonly used on websites like YouTube, Facebook, Gmail, and others. To test Flash performance, we used the Flash component of the GUIMark rendering benchmark.

Peacekeeper has been updated since we first started gathering notebook performance data, so we don't have results for all of the systems here. When running at full steam, the UL30A just trails our single-core Aspire Timeline config. That makes sense given that the Asus' processor is 100MHz slower.

Of course, the more interesting story is what happens when you enable the UL30A's Battery Saving mode. With its CPU clock speed capped, Peacekeeper performance drops by almost exactly a third.

The Aspire Timeline again proves a little bit faster than the UL30A—or a lot faster, depending on whether you have Asus' most aggressive power-saving mode enabled. In Battery Saving mode, the UL30A only just scores higher than an Eee PC.

Next up, we have 7-Zip's built-in benchmark, which tests file compression and decompression performance. We used the 32-bit client and let the test run up to 10 iterations.

Again, the Timeline's clock speed advantage delivers the win—but only just. Once more, the UL30A's Battery Saving config turns in a much lower score.

Video playback
Our next batch of tests highlights the UL30A's video playback performance. The chart below includes approximate CPU utilization percentages gleaned from the Vista Task Manager alongside subjective impressions of actual playback.

I used Windows Media Player to handle our DivX video, QuickTime for the others, and Firefox for our windowed YouTube HD test. QuickTime doesn't exploit the GMA 4500MHD's video decoding capabilities, so the CPU will be doing all the work there.

UL30A-A1 UL30A-A1 (Battery Saving)
CPU utilization Result CPU utilization Result
Star Trek QuickTime 480p 9-17% Perfect 32-77% Perfect
Star Trek QuickTime 720p 24-40% Perfect 57-100% Smooth
Hot Fuzz
QuickTime 1080p
30-73% Perfect 74-100% Some dropped frames, loss of AV sync*
Hot Fuzz
QuickTime 1080p (MPC-HC)
12-23% Perfect 26-37% Perfect
DivX PAL SD 8-17% Perfect 22-34% Perfect
YouTube HD windowed 43-57% Perfect 74-85% Perfect

In performance mode, playback was perfectly smooth for all five videos. CPU utilization was reasonably low, as well.

Switching to Battery Saving mode increased CPU utilization quite a bit, but that's to be expected given the associated clock speed drop. Playback was still perfect with our 480p QuickTime movie, our DivX BitTorrent download, and even with YouTube HD. 720p QuickTime content played back smoothly, too, although with the odd stutter here and there. 1080p playback proved more problematic, with noticeable dropped frames and the occasional loss of AV sync.

It seems a little unfair to make the UL30A's effectively underclocked processor handle all the number crunching associated with HD video decoding when the GMA 4500MHD is designed to handle such a task. Media Player Classic Home Cinema supports the Graphics Media Accelerator's video decode mojo, and as a free download, it's worth a shot. I primarily tested with MPC-HC in Battery Saving mode, but I observed much lower CPU utilization with our three QuickTime movies. Even at 1080p, CPU usage didn't exceed 37%. Playback was perfectly smooth at all three resolutions, as well.