Indeed, during a recent conference call, Jure Sola, chief executive officer of Sanmina-SCI, a large EMS company, told financial analysts: "There's no way in the world this industry can exist on the margins that we are delivering today".
Innovation is the route out of low-margin manufacturing. IBM, Xerox, AT&T and HP built their R&D capabilities with cash from unique products that commanded high margins - or, in the case of AT&T, from an outright monopoly. But those companies have a harder time justifying investments in research today. "It's difficult to make an ROI argument for creating fundamentally new scientific knowledge," says Mark Bernstein, president and centre director of PARC, the former Xerox think tank that was spun out into an independent subsidiary in 2002. "Faster product cycles and the increased focus on efficiency and productivity have made it harder for companies to have a long-term vision."
The loss of manufacturing and design could make it difficult for the traditional R&D powerhouses to innovate in the future. "Real breakthrough product development usually requires manufacturing and research to be located together," says NIST's Tassey. Supercomputers and high-tech weapons, for example, required close collaboration between engineers and manufacturers.
But PARC's Bernstein says R&D must become more global by necessity. "The breadth of research required to master a market these days is pretty significant," he says. "You're going to see a lot more partnering" around the globe to do research. Besides outsourcing manufacturing and design, many companies have opened their own dedicated R&D facilities in low-cost countries such as India and China. Innovation still occurs under the banner of a US corporation, but it happens elsewhere, employing lower-cost engineers.
To some observers, this may sound a death knell for the United States' current lead in technology innovation, but HP's Faber isn't overly worried. "I hate to sound like a Republican," he says, "but when I first came here 20 years ago, we had our own factories for sheet metal and screws and everyone thought we had to keep them. As we outsource, we just keep focusing on higher value-added work." HP's newer 64-bit servers are examples of products that are largely conceived and designed in the United States, he says.
It's clear, however, that this is a sensitive issue for the traditional R&D powerhouses. All but one (HP) declined to comment on the growing trend in outsourcing the "D" in R&D. At the same time, the EMS companies we spoke to denied they have any plans to expand into the "R" part of R&D or offer their own products for sale. Given their dependency on brand-name companies for business, it's unlikely that we'll see a "Quanta Labs" anytime soon. But as the EMS companies take on more and more design work and build up their engineering groups, there is little doubt that they will eventually have the capability to come up with their own ideas.
Quanta, for instance, has 1500 design engineers today. The plan is to expand that number to 7000 in the next couple of years. Fang thinks other EMS companies will follow suit. "The profits in manufacturing aren't large," he says. "So moving into design is an obvious choice. It's a natural evolution." Even if Quanta has no plans to sell its own products, surely one of those 7000 engineers will have a good idea up his or her sleeve.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.