For anyone interested in seeing what an early stage gorilla looks like, I suggest you listen to today’s WIND conference call. While revenues are not yet at tornado level, they are increasing, and evidence of their domination of the embedded systems grows stronger all the time. As they mentioned on the conference call, the trend to using a commercial operating system as opposed to an in-house system is increasing, and is driven by two powerful forces—the increasing complexity of embedded devices and the shorter development cycle, or what Wind calls “Time to Market”.
There are a couple of examples of the above from their conference call, and also from the recent news, that illustrate the above. First is the Palm win. Palm will now use Wind River's TCP/IP networking stack to provide Internet connectivity features for Palm OS® handheld devices. I think this is an excellent example of the first force moving business to Wind, increasing complexity. I think there is little doubt that technically Palm could develop their own networking stack. However, they made the decision that it made more sense to buy Wind’s than to build their own. They can now focus on the usability features, and services that the connectivity brings, rather than on the details of whatever the latest networking protocols will be a year from now. If Palm, an OS company made this decision, how much easier must it be for a non OS company to come to the same conclusion.
The next example that seems obvious to me is Microsoft, and their current problems with their set top box. If there ever was a captive market, it would seem to be all of the cable companies that MSFT has poured money into over the last few years. The only problem is that MSFT doesn’t have a product ready when its customers need it. Meanwhile, Liberate, Sony, Open TV, Acer, Hyundai, Pace and others, who mostly run their boxes on Wind software (some also work on other RTOS’s), are ready to sell to the customers that Microsoft counted as their own. If the greatest software company in the history of the world, with all of its resources can’t get the product to market in time, what chance does any other company have.
Both of the above are operating system companies. Sony is probably a much better illustration of the market forces that will benefit Wind in the coming years. They own Aperios, their in-house operating system. By all accounts I have seen, it is an excellent operating system, and I would bet that there are lots of engineers inside of Sony that think Sony should never use anything else (why should they spend all that money for run time royalties when the home grown solution will do). However, given the recently announced design wins by Wind, it appears that VxWorks can better handle the complexity (probably having to do with connectivity) of next generation products. There is no doubt that, given enough time, Sony’s engineers could solve whatever problems there may be with Aperios and connectivity, but meanwhile, the product must get out the door. Sony now has a set top box and a Network Walkman that use VxWorks that we know about. I’m sure that there are more products being developed that use VxWorks that we don’t know about. Once the momentum starts for outsourcing the operating system, how will Sony justify the cost of maintaining and constantly improving its in-house system? Personally, I think the writing is on the wall.
The advantages that Wind has go far beyond the above examples. They are forging alliances with the chip manufacturers that both increase the value of using Wind and make it easier for a customer to choose Wind. For example, they are the reference design for 9 of the 10 network processor manufacturers. What this means is that if a customer orders an evaluation copy of a new or prototype processor, he will receive the processor along with the reference software (VxWorks) for evaluation. In most situations, the easiest thing for the customer to do is to continue to use the same software for production.
Its not just with network processors. Wind is dominant in the areas of DSL modems, cable modems, bridges, routers, access multiprocessors, network gateways, and optical switches (they had 20 optical design wins last quarter). In short, anything where connectivity is an issue is likely to be dominated by Wind in the near future. The chip manufacturers know this, and that is why they all seem to be getting closer and closer to Wind all the time.
So why haven’t we seen tornado type increases in revenue yet? I think it is because so many of the markets Wind dominates are so new, and the royalty model that Wind uses tends to show revenues a quarter or so after the actual products ship. If you look at DSL or cable modems or set top boxes by themselves, the revenue that Wind generates may look like its in a tornado, however when you combine it with their other revenue, you don’t see the tornado.
For me, the key is that I just can’t see any credible competition for Wind. For the reasons I outlined above, I strongly believe that in-house operating systems will continue to be used on the more simple products, but that they will be used less and less on cutting edge networked products. With the value chain Wind has developed (hundreds of companies trying to jointly develop seamless solutions), I don’t think Microsoft can overtake them, especially given the legal environment they currently operate in. Palm is now a customer, so I don’t see them as a real threat. I see Linux as less of a threat than I did several months ago. Linux does not provide a total solution, and I doubt that it will ever provide a complete solution that is free. As Wind said in their conference call today, what their customers are looking for is a complete, one-stop solution. They do not want to use a kernal from one source, a browser from another, and a networking stack from a third, and have to cobble them together themselves. There is no longer time to do that.
I also have a guess as to the future. I think that the cell phone manufacturers will find themselves in a situation similar to Sony and Palm when they start to deal with high-speed data transfer and not just voice. I bet that at least some of them will end up with Wind. Motorola selected Wind River software for its advanced-feature phone currently under development, and I’m sure others will follow.
All of the above is purely my opinion. |
