WIND RIVER SYSTEMS
PROJECT HUNT REPORT
OVERVIEW
Wind River Systems (WIND) is the leading provider of real time operating systems and related tools and consulting services for embedded computers. An embedded computer is a microprocessor that is incorporated into a larger device and is dedicated to responding to external events by performing specific tasks quickly, predictably and reliably. Over 90% of the processors in use today are in embedded systems. Historically, most companies have written their own operating systems for their embedded computing needs. However, as microprocessors have become more complex, and as the number of different microprocessor families has multiplied, companies have more and more turned to commercial off the shelf (COTS) systems.
WIND estimates that they have an approximately 35% share of the COTS market. The number two company in the industry, Integrated Systems (INTS) has a market share slightly less than WIND. On February 15, 2000, WIND expects to complete its acquisition of INTS. Combined, they will have a two thirds market share. WIND estimates that all of the COTS vendors combined currently have about 15% of the total market for embedded systems; the other 85% is from in-house systems. WIND's business model is to charge for an up front development license, plus a royalty for each copy of the operating system used in a customer's products.
WIND was founded in 1983 and is headquartered in Alameda, California. INTS was founded in 1980 and is headquartered in Sunnyvale, California. Each company has approximately 600 employees.
MARKET DESCRIPTION
Customers:
Companies building products containing microchips in such industries as automotive, aerospace, office automation, telecommunications, consumer electronics, etc. Wind River's customers include Boeing Company, Cisco Systems, Inc., Ericsson Radio Systems AB, General Motors Corporation, Hewlett-Packard Company, Hitachi, Ltd., Hughes Aircraft Company, Lucent Technologies Inc., Intel Corporation, Lockheed-Martin Corporation, McDonnell Douglas Corporation, Mitsubishi Electric Corporation,
Motorola, Inc., Network Computer, Inc., Nippon Electric Corporation, Northern
Telecom Ltd., Raytheon Company, Siemens AG, Sun Microsystems, Inc., TRW Inc. and many others.
Competitors:
The primary competition for WIND comes from companies that develop real time operating systems in house rather than purchase them from independent software vendors. Competitors other than in house software departments include Accelerated Technology, Inc., Mentor Graphics, Inc., Microsoft Corporation, Microware Systems Corporation, QNX Software Systems, Ltd., Sun Microsystems, Inc.; and Symbian Inc.
FINANCIAL AND STOCK INFORMATION
Please see the SI analysis at siliconinvestor.com
I have not included it here due to formatting problems. WIND is growing its revenues at approximately 40% per year, has gross margins in excess of 90% and operating margins of around 25%.
As of January 27, 2000, WIND had a market cap of $1.366B and INTS had a market cap of $.715B. Combined revenues were $171 million, $214 million, and $265 million for the years ended January 31, 1997, 1998 and 1999 respectively, and $191 million and $229 million for the nine months ended October 31, 1998 and 1999.
G/K CHARACTERISTICS
Is there a discontinuous innovation or a proprietary open architecture?
Yes. I believe that the idea of outsourcing maintenance and development of an RTOS is a discontinuous innovation. When a company no longer has to maintain their internally developed systems, they gain time to market advantages. Additionally, they are free to switch processor families (WIND supports approximately 35 microprocessor families; far more than any other vendor) without having to port their systems. WIND has over 400 partners that integrate their software tools with VxWorks, WIND's proprietary operating system, and there are over 40,000 software developers using WIND tools.
Does it have the potential to grow into a mass market phenomenon, become a standard?
Again, I believe the answer is yes, and I believe VxWorks is already almost a standard. My(unscientific) review of help wanted listings for embedded engineers showed that where knowledge of a specific RTOS was not desired, the requirements would contain a statement such as "knowledge of an RTOS such as VxWorks, pSOS, or QNX.", and nearly always listed in that order. pSOS, Integrated Systems' product, will be owned by WIND in a few weeks. QNX, the number 3 RTOS, is privately held. As the FM indicates, enabling technologies such as operating systems are often a winner take all game.
Are there high barriers to entry and high switching costs?
Again, yes. Mission critical software used in items such as anti-lock brakes and pacemakers simply will not be used until it has been around for a long time. Additionally, engineers, like most other people, like to work in a familiar environment. This is probably one of the biggest reasons that 85% of the market is still in house. Many of these departments will not change until they are in pain (such as not having their in house RTOS ported to a new microprocessor in time for a new product launch, or not having the required communication protocols available. Imagine the complexity of going from an 8 bit system to a 16 bit system to 32 or 64 bits). At some point, the intellectual property put into the RTOS must be amortized over more than just one company's products, even if the company is a GM, Cisco, Sony, Motorola or Texas Instruments. In fact, all of these companies currently use a mix of in house and commercial programs. When they do decide to standardize on a commercial program, they are more likely to standardize on the company with two thirds of the market than anyone else.
Have value chains developed, and have they crossed the chasm?
Yes and yes. I believe that WIND is currently in the bowling alley. The easiest way to look at the value chains is to look at some of the pins in the bowling alley, and the alliances WIND has developed to provide complete products and knock over the pins:
Wind River, Intel Validate Intelligent I/O Architecture With Real Time OS for I20 and Intel Integrated RAID Building Blocks
biz.yahoo.com
Lucent Technologies, Wind River, Trillium Digital Systems Leverage Expertise to Reduce Time-to-Market for Internet Phones by as Much as Six Months
Integrated Hardware/Software Development Kit Creates Complete Solution for Voice Over Internet Protocol Phones
biz.yahoo.com
FlashPoint, Wind River Introduce First Application Software Development Kit for Digita-Enabled Cameras
wrs.com
Wind River?s Tornado and VxWorks Selected for Cisco NetWorks Program
wrs.com
There are many more similar examples. 34 cable modem vendors and 29 DSL equipment vendors have selected WIND. WIND claims more java design wins than all other vendors combined (before the acquisition of INTS. After, it is likely to be more like 75-80%). For a great overview of the value chains that have been developed, I recommend a corporate presentation that can be found at media.corporate-ir.net
Existence of hypergrowth?
Not yet. Revenues are growing at about 40% per year, so they definitely are not in a tornado yet.
SUMMARY AND ANALYSIS
With their design wins in things like cable modems, servers, NAS, SAN, I2O, Infiniband, etc. that have mostly not yet started to ship in quantity, it is possible that they could go into a tornado in the near future. Personally, the indication of a tornado that I am looking for is for a major, multi product company to announce that they have standardized on Wind River Systems' products. Once that happens, I believe many other companies could fall into line very quickly. I am hoping that the merger of the two largest companies in the industry (I have read that no one else has even a 5% market share, and that MSFT is currently less than 1%) will prove to be the type of catalyst for WIND that the Ericsson deal was for QCOM.
According to the FM, it is still too early to invest in an enabling technology when it is still in the bowling alley; it is not until it has entered a tornado that one should invest. Because of the reasonable valuation placed on WIND by the market compared to its earnings and revenues, I have made an exception for my portfolio. I feel that, with the design wins they already have in future mass markets combined with their business model of charging a royalty for each copy of their software (i.e. for each digital camera, printer, cable or dsl modem, raid/I2O/Infiniband storage system) strong future growth is almost assured.
The primary risks I see are Microsoft and Linux. Microsoft is only a threat if they buy QNX, in my opinion. Even if they did buy QNX, WIND would still have a huge lead, both in market share and in number of processors supported. However, with MSFT's immense resources, that lead would be significantly offset. Very few companies can afford to give AT&T $5 billion to guarantee that half of their set top boxes will contain CE. My bet is that a significant portion of their remaining boxes will have VxWorks. In fact, one of WIND's big advantages is that it is not MSFT.
Linux, is also a significant threat. My current feeling is that they have a long way to go to get to where WIND is, they have a tremendous amount of momentum behind them, however it will be at least several years before the same types of value chains develop around embedded linux that are currently already developed around VxWorks. If linux ultimately does become successful, my feeling is that the RTOS market will then become a royalty game instead of a gorilla game, and that WIND will either be a king or a strong prince.
RESOURCES
The SI thread at Subject 2113 has a great index at Subject 31747
comp.os.vxworks is a newsgroup for people developing products using VxWorks. I'm not an engineer, but it is interesting to go through the messages once in a while to see some of the companies whose engineers are participating in the discussions.
WIND's home page at www.wrs.com has a lot of information
Monster.com at jobsearch.monster.com shows 611 job listings as of this writing. It is another source of which companies are using VxWorks (for competitive reasons, companies often don't want the competition to know what they are working on, so it is often difficult to tell which RTOS they are using). |
