Portege R500. How did they get the 7 hours battery life?

Fancy a 7 hour UMPC? Well the Toshiba Portege is said to have a 7 hours battery life in 800gms of weight. Why can’t UMPCs use the same technology? Wouldn’t it be great to have an Intel Core 2 Duo 1.2Ghz GMA950 ultra mobile PC using the new U7600 Intel 65nm processor? Wooosh! Well, let me bring you some news. Intel based UMPCs use exactly the same technology and are often MORE efficient than the Toshiba Portege R500. 7 hours into a ultra mobile PC just doesn’t fit.

As usual with manufacturers battery life claims, its mostly smoke and mirrors. Firstly, Toshiba have used a nice testing trick that we’ve seen before. The 7 hours life is based on the JEITA test. To be blunt, its rubbish! I analyzed that test a few months back when I was suspicious of Fujistsu’s 4 hours ultra mobile PC battery life claim and found that its one of the easiest battery tests you could ever give a PC. Its a hands-off low-load and idle test and is effectively useless in determining real-world figures. Real-world usage with WiFi, reasonable screen brightness, and rich browsing is surprisingly heavyweight and easily knocks off about 30% of battery life from the JEITA test so in this 7-hour case, you’re already down to under 5 hours. Secondly, we’re dealing with an Intel processor and GPU here. Devices based on these chips can idle nicely down to very low power drain levels but can also burst to very high drain levels. A 1.2Ghz Intel CPU and GPU running at full tilt runs at least twice the drain of say, the AMD Geode LX800. Its not surprising that some R500 testers are reporting around 4 hours battery life because Vista and media-rich websites will use all the processing power they can get to perform their tasks and when the task is done, the user will just keep piling on new tasks! The final important point to note is that there a huge battery on the R500. In fact its about double the size of a standard ultra mobile PC battery at around 60W/hr. The same marketing trick is used with the Flipstart UMPC. It ships with a relatively massive 40W/hr battery and appears to have a good battery life. When you analyze the figures you can see why! Oh, there’s one other thing too. The 800gms weight is for an SSD version of the R500. I’m guessing that the JEITA tests where done using the SSD version which would have saved even more power.

If you factor all these elements in and scale the whole Portege R500 product down to ultra mobile PC sizes, weights and heat dispersion characteristics, you end up with a 30W/hr battery and around 2 hours of useable battery life which is exactly what we had 1 year ago. The only way you could scale the Portege down into a competitive UMPC-sized package would be to limit the ‘burst’ capability of the silicon and simply put, to slow the user down. I wonder how you could do that? Cap the clockrate perhaps? Well well well! That’s exactly where we are today with the Intel-based Stealey UMPCs.  Menlow, due at the end of 2007, will be interesting. There’s some new and quite mind-boggling chemistry and physics going into the 45nm ‘low-k dielectric’ that should help to bring the watt-per-clock and thermal characteristics down a lot but don’t expect miracles. Bandwidth capping will remain an important battery, heat-saving and user-slowing technique.

So does the R500 look any good after that? Well, I guess so. Its using ultra-efficient components and that’s always good in my book. Its got a nice transreflective, LED back-lit display and if you need a full sized keyboard and DVD burner, its there! Apart from that though, I can’t really tell you much about these notebook things. I’ve never owned one and probably never will.