On Banias, Centrino, and on benefits and premiums

Looking over the reviews of those brand-new Centrino-based notebooks one could get really enthusiastic. However, after some second thoughts there is not so much to be excited about.

1. Starting off with what intuitively is considered as the most important part: Banias-CPU. Impressive specs in respect to TDP and performance, no doubt. However, comparing the average power-consumption specs in mobile use to those of former products (P3 or PIV) there is barely a difference. That is so because CPU accounts only for some 10% of total power-consumption in mobile computing (for any user-model used for battery-benchmarking) which is 10-20 Watt, so we are talking about 1-2 Watt for CPU. What is left from the glory after a second thought is a saving of nothing more than one Watt.

I hear you saying this is not only counterintuitive, but obviously nonsense, as every review has shown some 50 percent increase of battery-runtime? Right, except for this nonsense:

2. Centrino is more than just Banias: It includes a platform including chipset and grafics. So we take a closer look on the chipset first: We are told Speed-StepIII is capable of making mobile computing much more efficient in terms of power-use. Surprisingly, evaluating this one you find barely any difference for battery-runtimes using minimal, variable and maximal settings of Speed-StepIII. Surprisingly? Not really, as power-consumption of CPU is pretty irrelevant for these benchmarks (see above). So what is left on the platform-side: Optimizations of the power-supply? Lower voltages for the chipset? Next-generation grafics-chips? Yeah, sure: All of it, and more, a little bit here and there - taken together accounting for more energetic benefits than from Banias. Maybe we have some better clue now why Centrino may only be called a Centrino if CPU and Platform are used together. But by far not enough to really explain the increases in battery-runtimes completely.

Now, if this still is not all the reasons for the impressive increases of runtime, what else could it be?

3. A notebook is more than Centrino: Two (not so) trivial things are added: A display, which accounts for one third of the power-consumption alone. Plus, battery-packs. Without being able to quantify the power-saving of the displays used exactly, it is safe to say that displays of the same size and resolution consume approximately 10% less energy to operate from one generation to the next; for batteries energy-density increases by some nine percent per year. So there is significant benefit coming from these components - attributed to Centrino, claiming not only the merits of, but trying to capitalize on the premiums for these benefits by means of Centrino-pricetags.

As for the most careful conclusion insofar, the benefits delivered from Centrino-based notebooks are delivered not exactly as we expected, as some 300 million greenbacks try to convince us to accept the premiums asked for Centrino-brand Notebooks.

To conclude, this really turns out to be nothing much more than another chapter of the Book of Marchitecture, I admit to be still sort of impressed - about how smart this has all been invented, using the known intuitive dimensional misjudgements discussed above.

Klaus Fehrle
CEO f.a.l.k. marketing GmBH, Filderstadt