Dave:
Just thought I would add my two cents, hoping to clarify something.
<<The $1.50 number is a guess based on some analyst saying he thought WIND would get between $1 and $3 per chip... Obviously, we have no idea if the I2O royalty decreases as volume increases.>>
It is my understanding that the Intel deal carries a flat royalty payment, regardless of total volume shipped. This same issue came up for me a while back, when Intel announced the i960 RD chip. At the time, I was concerned that WIND's royalty stream would not be intact on future generations of RISC processors, such as the RD. But WIND does continue to collect the royalty. My conversations with WIND management have led me to belive that this royalty is flat for as long as the IXworks OS is used, including migration to successive generations of chips. Also, WIND has indicated a few times on various conference calls that they have several other large application royalty deals in the works, and are waiting to announce them (see Allen's post about this subject. Sorry, I forget the number). Such statements also lead me to believe that the Intel royalty is a flat number, regardless of volume.
As for the actual number, we have all speculated. This thread seems to have settled on $1.50, since WIND once mentioned that the royalty was between $1.00 - $3.00, with the number being closer to the $1.00 level. Beyond that, I don't think we can reliably tighten our estimate any further, but $1.50 seems to be as good as any number.
As for the potential number of I20 chips, the question is still up in the air. Like you, I would love to believe that i20 will eventually migrate to the desktop.
<<As I said in my post to Mark, there are a number of reasons to believe that I2O will end up in every PC. The foremost reason is that Intel owns the PC hardware platform, and will, therefore, push those technologies that increase its profits.>>
Your logic is very sound here, but for me, it is still to early to tell. When I make estimates, I limit the number of chips to Intel servers, since this is what the chips were initially designed for, and they need to be sold there before any migration to the desktop becomes valid. Also, not being an engineer, I have no way of knowing if i20 on the desktop carries enough benefits to the consumer in order to warrant the extra expense. Perhaps you could further elaborate on exactly what i20 will do on the desktop, so that I could be more comfortable including these numbers in my forecasts.
In the meantime, even the server business should be good enough to make us quite wealthy, given the world's growing appetite for faster, cleaner servers. As you indicate, most servers should need multiple i20 chips, increasing the value of this royalty stream. So much the better for all the longterm WIND shareholders.
Hope this is helpful.
Regards,
Jason Cogan |
