9. Testing: Methods, Benchmarks And Systems

For this set of tests I used selected benchmarks from BAPCo's MobileMark 2005 and FutureMark's PCMark05 test suites.

MobileMark 2005 uses real world office applications to measure battery life under different conditions. It also provides a useful measure of the performance of the combined components in a mobile computer that is running on battery power. For these tests I used the MobileMark 2005 Productivity, Wireless Browsing and Reader benchmarks.

PCMark05 produces performance scores for certain discrete components in a computer while the computer is running on AC power. For the tests conducted here I used PCMark05's CPU, memory and hard disk drive benchmarks.

For more on how we use the MobileMark 2005 and PCMark05 benchmarks see How We Test Laptop and Notebook Computers for Home and Office.

For comparison I tested two ultra lightweight laptops that are very similar to the OQO Model 01+ in some ways. These two computers achieved very similar results on all my tests so I averaged the results for each test and present the average together with the OQO's results on each test. Here are the basic specs for the Model 01+ and the two laptops.

OQO Model 01+: 1 GHz Transmeta Crusoe TM5800 512 kB L2-Cache; 512 MB 266 MHz DDR memory; 30 GB Toshiba MK3006GAL (P-ATA, 4200 rpm, 2 MB cache, 15 ms average access time); Li-Polymer 2 cell 14.8 Wh battery; Windows XP Pro SP2

Toshiba U100-S213: 1.2 GHz Pentium M 753 ULV CPU 2 MB L2-Cache; 512 MB 333 MHz DDR memory; 60 GB Toshiba MK6006GAH (P-ATA, 4200 rpm, 2 MB cache, 15 ms average access time); Li-Ion 3 cell 28 Wh battery; Windows XP Pro SP2

Fujitsu Lifebook P1510D: 1.2 GHz Pentium M 753 ULV CPU 2 MB L2-Cache; 512 MB 333 MHz DDR memory; 30 GB Toshiba MK3006GAL (P-ATA, 4200 rpm, 2 MB cache, 15 ms average access time); Li-Ion 6 cell 35 Wh battery; Windows XP Pro SP2

The MobileMark 2005 Tests: Preliminary Considerations

You'll remember from earlier discussions that the maximum brightness of the OQO Model 01+ display is around 36 candelas/meter^2. That's as bright as the display can be set. The MobileMark 2005 tests require a display brightness setting between 60 and 70 candelas/meter^2 in order to assure the consistency of tests and results. For the MobileMark 2005 benchmark tests I set the OQO's display to as bright a level as possible, 36 candelas/meter^2. I tested the Toshiba and Fujitsu at regulation MobileMark 2005 levels of 60 to 70 candelas/meter^2. Display brightness is the greatest battery life thief there is. So the differences in brightness between the Toshiba and Fujitsu and the OQO put the two laptops at a battery life disadvantage. They would have run longer at a setting near 36 candelas/meter^2, probably as much as 60% to 75% longer.

Also the OQO's battery has half to much less than half the capacity of the batteries in the two laptops. With its optional double capacity battery, the OQO would have run about twice as long as it did.

So, though the test results may be imperfect, as you'll see, they're still quite interesting.

Note: All of the battery-based tests discussed here were done with Microsoft's USB 2.0 battery drain patch for Windows XP SP2 installed. For more on the patch that applies to all computers running Win XP SP2 see these articles: