Electronics manufacturing generally relies on scale and automation, with machines reproducing tiny features across millions of identical devices.
Some kinds of work, though, still rely on a human touch.
In an industrial park in Beaverton, one of Oregon's biggest technology companies makes handcrafted electronics. Cascade Microtech employs former auto shop workers and circuit board assemblers and others to customize testing equipment for all the world's biggest semiconductor manufacturers.
The tools they produce help those chipmakers evaluate new designs and test future products, and every one of those clients wants something different. So Cascade Microtech's employees customize wafer probes to within the width of a human hair, tuning them to specific radio frequencies to meet each client's needs.
"While we have a standard product in a standard catalog, the reality is very few of those go out in a standard format," said Cascade Microtech chief executive Michael Burger.
Thirty years after its founding, and 10 years after its initial public offering, Cascade Microtech is finally thriving as the electronics industry races toward wireless technology, fueling demand for the Beaverton company's tools.
Revenues hit a record $120 million last year and were up 16 percent through the first half of this year. Net income doubled last year, topping $13 million and the company boasts a gross profit margin north of 50 percent.
And yet Cascade Microtech remains a small player in a massive industry, subject to the whims of clients whose revenues run into the tens of billions of dollars.
Cascade Microtech
Founded
: 1984, by Tektronix engineers who licensed their wafer probe technology from Tek.
Business
: Equipment for designing and testing semiconductors, especially chips with wireless capabilities. Customers include all the largest chipmakers.
Employees
: 448, including 300 at its Beaverton headquarters.
Revenue
: $120 million in 2013, up 6.2 percent.
Profits
: $13.4 million, up from $6.1 million.
"It puts a lot of pressure on companies our size to show we are relevant," Burger said.
Like many of Oregon's established technology companies, Cascade Microtech traces its roots to Tektronix. Founders Eric Strid and Reed Gleason licensed technology from Tek, their employer, and then set out on their own in 1985.
Cascade Microtech went public in 2004 in what turned out to be the last Oregon tech IPO for at least a decade (none have come since, and none are in the pipeline).
In hindsight, Cascade appears to have jumped the gun. The company had sales of just $64 million in the year it went public, well below the $100 million threshold investors typically expect. Struggling for scale, the company lost $58 million during a four-year stretch beginning in 2008, with shares sinking near $2.
Recovery began in 2010, when Cascade bought its chief rival, the testing division of a German company called SUSS MicroTec.
Cascade hired Burger the same year, and the combined company increased 80 percent in size while it cut operating costs by consolidating factories, product lines and sales channels. The company introduced new products with higher profit margins and bought a handful of smaller companies that immediately added to Cascade's profits.
The stock is now trading around $10, giving Cascade a market value of roughly $175 million.
The semiconductor industry is notoriously cyclical, which makes life brutal for many small suppliers. Cascade benefits because it serves chipmakers' engineering departments, rather than their manufacturing operations, so it's less vulnerable to market swings than other toolmakers.
And Cascade's historical specialization in radio frequency (RF) technology has benefitted the company during the ongoing mobile boom.
"Anything associated with consumer or handheld devices is doing very, very well," Burger said.
That RF niche has also positioned the company for the emerging fervor for the "Internet of Things," an industry term that describes wearable technology, connected appliances and other everyday objects connected to the Internet by wireless chips.
Cascade tools can help design all those new products. Amid the unveiling of the Apple Watch this week, fitness trackers and other new wireless technology, Burger said the new products create a big new opportunity for his company.
Provided Cascade can stay ahead of its customers.
That requires Cascade to shift from a Tektronix-style engineering skunkworks operation, Burger said, into a business so closely attuned to its customers' needs that it knows what they want before they do.
"We're going to accelerate our R&D spending... because we need to keep up," Burger said.
Cascade owes a lot of its current success to its fortune niche in the thriving wireless space, according to Brett Piira, an investment analyst who follows Cascade for B. Riley & Co. But the company is standing out by delivering solid financial results, strong enough that it can afford to spend more on research as a down payment on the future.
"It's showing that they're not really satisfied with just relying on where they are," Piira said. "They're going to keep innovating."
Piira rates Cascade a "Buy" with a $14 price target.
Before Burger ran Cascade Microtech he was a division president at Flextronics, a big contract manufacturer. There, production was all about volume and replication.
Cascade's success relies on human touch and customization, and a stroll through the factory floor of its Beaverton headquarters finds employees at lab benches, peering through microscopes refining wafer probe tips. Across the hall, workers are customizing a million-dollar probing system to meet a client's specifications. In a separate clean room, Cascade Microtech has invented its own process for creating microscopic probe tips for clients working on especially tiny products.
Chipmakers such as Intel can produce microprocessors with features just a few atoms across. The scale Cascade works in is small, but not nearly that tiny.
Whereas Intel will make a million chips a day with exactly the same dimensions, though, Cascade's customers demand variety from every tool. That's the kind of individualized work, Burger said, that no machine can reliably produce.
"We haven't been able to find the tools," he said, "that can do it again and again."
-- Mike Rogoway; twitter: @rogoway; 503-294-7699