3. Mobilemark 2005: Conclusions

The VGN-TX1XP landed in second place in the office productivity performance test. In theory, it should have been exactly as fast as its sibling, the VGN-T350P. Its somewhat slower performance stems at least in part from the well-known observation that files never reside in exactly the same spots on disk on different systems, nor do disk partitions end up being laid out in exactly the same way. The location of files and partitions can, however, have a decided influence on hard disk as well as total system performance, because the data transfer rate of a hard disk is a function of disk platter radius. Another factor is that the VGN-TX1XP uses a smaller 1.8" diameter hard disk, which has lower performance charactistics than the 2.5" drives sometimes used in lightweight laptops.

In addition, performance metrics (like all measurements), are always subject to a margin of error of two to three percent. In other words, when MobileMark 2005 measures small differences between devices, this is really more of academic interest than of practical significance. No user can discriminate among differences in system response times that may be measurable only in hundredths or tenths of a second.

What users will definitely notice is what happens when the CPU clock is fixed at 600 MHz as when running the VAIO "Maximum Battery" power schema. This causes performance values to drop to 117 points. Adding 512 MB of RAM seems to make no real difference in performance, either: given the margin of error, 120 points means the same thing as 117 points.

Of course, the payoff is in the area of battery lifetime. Use of the "Maximum Battery" power schema extends battery life by 46 minutes as compared to the more typical Windows "Portable/Laptop" power schema. Giving up on an added SO-DIMM and using only the hard-soldered 512 MB of RAM on the motherboard adds another 13 minutes to battery lifetime, and extends that total period to nearly 7 hours!

Running the less demanding read and search test produces total battery lifetimes of 7.5 hours, if you remove the SO-DIMM memory module and set the fan to Silent Mode. This slows the CPU clock rate down to a mere 240 MHz, because the system slows the FSB to a paltry 40 MHz. Even using the Windows "Standard" power schema delivers battery lifetime of about 6.5 hours.

What does this teach us? Anybody who truly wants to maximize battery lifetime on at least an ultra-portable laptop should use only the hard-soldered RAM, forgoing plug-in memory modules!