I barely got the new netbook unpacked and plugged in before emails started arriving asking “where are the benchmarks?” Actually, for folks curious about the slightly tweaked Intel Atom N280 CPU (s intc) over the pretty standard N270, it’s a valid question. Just remember that we’re talking about some very minor differences in the architecture here: the N280 supports a faster front-side bus speed over the N270 — 667MHz vs 533MHz — and bumps the clock speed from 1.6GHz to 1.66GHz. That’s hardly a major change, so I wouldn’t expect a major difference.
The most fair comparison I can do is between the MSI Wind and the Toshiba NB205. When I ran the CrystalMark tests against the Wind, it had 1GB of RAM just like my NB205 currently does. I’m likely to upgrade that in the near future, but this makes for a good head to head comparison.
My pre-expectations were generally met with the Toshiba NB205 earning an overall score of 29,309. That compares to the MSI Wind’s score of 27,365 from last year, or about a 7% increase. Both have a 5400 RPM hard drive, but the NB205 earned 7,656 marks compared to the Wind’s 6,780. The difference is due to the faster read and write speeds on the Toshiba drive — they were in the 54 to 57MBps range for sequential reads and writes. The hard drive in the Wind could only push sequential data at around 42 to 44 MBps. Random reads and writes are important too, due to disk fragmentation, and the Toshiba’s drive edged out the one in the Wind there as well.
After I ran the CrystalMark program the first time, I thought to run it again. Why? Because I was using the Toshiba Power Saving program in “Full Power” mode and it’s actually unlikely that I’d be doing that while mobile. It’s also good to see what the minimum performance on the netbook is. To this, I unplugged the device from the AC adapter. I have the Toshiba Power Saving app configured to run in “Long Life” mode whenever it’s on battery power, so pulling the power cord immediately changed the power settings. Mostly, they dim the backlighting and reduce the amount of time before sleep mode, etc… but the “Long Life” mode also scales back the CPU to save some juice as well. That’s why I decided to re-run the benchmarks — to see what impact there was on the overall system.
The results were telling: an overall score of 21,625 marks. That’s nearly a 26% drop in overall benchmarked performance. Everything from graphics and hard drive to calculations and memory dropped as less power was provided to the components. Good to know when you’re on the road with a netbook and running on battery power. It should also provide some information on how to balance the compromise between performance and battery life with a mobile device.