The board business is changing dramatically, not just technologically, but fundamentally. By "board business," I mean off-the-shelf circuit boards meeting PCI, CompactPCI, and VME specifications and standards.
In the 1990s, the primary board markets were industrial controls, military/COTS, and telecom, with small markets for boards in medical, transportation, and some scientific applications in particle physics. When PC motherboards and PCI slotcards became prolific and cheap, many of the market segments for traditional 19-in. racked boards and boxes changed dramatically. Most industrial and medical applications went to some variation of a PC. The three primary markets today for board-level products are military/COTS, telecom/communications, and embedded PCs in industrial applications. Herein lies the problem.
During the Great Telecom Bubble, including the Dot-Com Bubble, numerous board makers dropped their traditional products and markets like MIL/COTS and industrial and dove headlong into telecom, many with PC-based technologies. The bubbles burst, and we saw an average drop of 65% in boards shipped to telecom companies over the last three years. Telecom had become the heroin of the board industry, and many companies suffered tremendously.
While the telecom debacle was unfolding, some companies concentrated on PCs in industrial controls, and others focused on PowerPC processors in the MIL/COTS markets. The industrial markets grow at about 7% in a really good year when the world economies are running on all cylinders. So the growth prospects are small, but the industrial controls market is very large and diverse. We didn't see gargantuan drops in shipments to industrial applications.
The demand for boards in military and COTS markets started growing 15% or more a year in the mid-1990s. With the latest developments in the war on terrorism, they're now growing at about 20% per year. While 85% to 95% of CompactPCI board sales go into telecom, about 45% of VME boards now go to MIL/COTS applications. VME board shipments to MIL/COTS applications this year alone are almost as large as the CompactPCI shipments to all applications worldwide.
In telecom applications, users just want cheap commodity hardware, heavily customized, for their gateways, routers, and protocol boxes. The only value you can add in that market is manufacturing value. The world rewards manufacturing value at about 8% gross profit margin (GPM). And, shipments to telecom are low volume.
Typical GPM of boards shipped into MIL/COTS applications run above 45% because the primary value added is intellectual value. Unlike PCs, these boards must run in extreme environments, and many of them are conduction-cooled.
The board markets today show only three major segments: low-volume/low-margin customized commodity products for telecom; low-volume/high-margin customized products for MIL/COTS and other extreme applications environments; and ruggedized commodity-oriented PCs for industrial applications.
As this market transition continues, we'll see hefty consolidation of board suppliers in 2004. In fact, it started in 2003. Along with their higher GPM, MIL/COTS boards have much longer life cycles, typically five to 10 years. So, MIL/COTS companies are selling for 1.5× to 2.0× sales. But telecom-focused companies are seeing much lower valuations, more like 0.5× to 0.8× sales. Blame their lower GPMs and product life cycles measured in months.
The MIL/COTS markets will grow at 15% to 20% as the military inserts new technology (i.e., faster processors) into legacy systems. About 80% of the MIL/COTS board shipments for the next few years will be new VME boards for existing VME systems. Only about 20% will be for new systems. As for telecom, market researchers don't see any significant improvement in equipment sales until 2010. So, telecom-focused board makers will shift their heroin-telecom addiction to the crack cocaine of networking applications. In 2004, we'll see how many board makers can survive the transition, if there's a real market in networking, and if they go through another severe withdrawal. It's crystal clear that the secret of success is diversification, but the hallucinogenic effects of telecom and networking have many in their addictive grip.