I can't remember how long I've been putting out that list of licensees, but early on, I indicated that it is a flawed method going based only on revenues.

However, even though it's flawed, I can't come up with a better method without more info from the company. If anyone can, please do! Also, even though it IS flawed, there is some basis for merit in using it. Mainly, that lies in the idea that MOST of the companies that have signed are predomminately electronics companies. Therefore, it stands to reason that most of the products they sell contain microprocessors. With that in mind, it stands to reason that the revenues that they generate reflect in a direct way to how many microprocessors they sold. And on AVERAGE, blending 57 companies MIGHT produce a refletive metric.

All that being said, obviously, Fujitsu's products might contain a lot more microprocesors than Asustek's. Conversly, Fujitsu may charge a lot more for their products and have higher profit margins which would in turn affect how accurate revenues would be in reflecting the amount of microprocessors sold in their products as compared to Asustek, etc. etc. So all in all it IS admittedly a very "loose" way of analyzing the licensing. It's possible that AVERAGING lessens the inaccuracy but certainly not necessarily. There are obviously other variables including overall market conditions, when the deal was cut, the financial health of a particular infringer, etc. etc. etc.

Ultimatley, though, it seems to have provided at least a range of probable license values based on revenues that quarter by quarter has ranged as high as $1 per every $960 in revenue, to as low as $1 per every $5452 in revenue. That is UNTIL the Asustek license last quarter, which equates to $1 for every $29793 in revenue. An order of magnitude shift of about 1/10th previous fees. I don't know a bunch about Asustek, but I find it hard to believe that revenues are based on products that contain 1/10th the amount of infringing tech than the other 56 companies. In my opinion, the list also has some value in pointing out those kinds of anomolies.

I'm going to repost the COMPLETE list in my next post just for those who might be intersted in the data.