Applied Microsystems Corporation (AMC) agreed last week to sell its embedded development technologies to Metrowerks, a software development company owned by Motorola. The sale marked the end of a long process driven by forces affecting the entire development tools industry and cemented by the bursting of the new economy bubble.
AMC started off in the late seventies as an engineering consulting company. After designing an emulator for a contract, the company went into the development tool business, with the recently designed emulator as its core product. In the ensuing years, it became a major player in that market.
Problems arose when semiconductor vendors collectively put run-control capabilities on chip, making a significant portion of a standard in-circuit emulator's capabilities redundant. As more and more debug capabilities were implemented in silicon, the value of the in-circuit emulator declined.
According to Arnold Berger, who was AMC's director of research and development from 1996 to 2000, this was part of a trend affecting the entire development tool industry.
“All of the traditional development tool companies are facing the same issue,” said Berger. “It used to be that the semiconductor companies had divisions that supported their development tools, because that was the only way they could do it. Then the pendulum swung the other way and the semiconductor vendors worked with third party companies, like Hewlett Packard and AMC, who provided the tools. The semiconductor companies were like a benign uncle that would occasionally give you money. Now they're finding they have to bring the tools in house again because by putting all this stuff on chip they're destroying the value proposition of their third parties. So they have to buy development companies. Motorola bought Metrowerks. Intel bought Cad-UL. I think it's going to have to stay that way, because everything is going on chip.”
Aware of the implications of this trend, AMC made several attempts to broaden its offerings. The principal product in these efforts was CodeTEST, which shifted AMC's customer focus from hardware and firmware developers to embedded software application developers.
CodeTEST met with some success but not enough to change the company's fortunes. The same was true of AMC's foray into development tools for console game systems through contracts with Microsoft (XBOX) and Nintendo. While successful, they didn't make enough money to offset the company's losses in other areas.
The customers supporting AMC's product expansions were large telecom and networking vendors, such as Nortel and Cisco. The last straw came when these market sectors collapsed in 2001. AMC experienced a precipitous decline in sales and, even after several rounds of layoffs and other cost reduction measures, it no longer had the resources to invest in new offerings. AMC was losing money, and eating deeply into its coffers.
After failing in several attempts to combine with other companies to preserve what remained of its value, AMC decided that its only viable course of action was to sell its assets, which is how the deal with Metrowerks came about.
Metrowerks agreed to purchase the majority of AMC's development portfolio for around $3.9 million and plans to continue selling most AMC products. The company says it will also offer support for existing AMC customers, as well as those who purchase tools between now and the realization of the sale. Metrowerks plans to extend offers to a number of AMC employees and also to train its own employees to work with AMC technologies.
Applied Microsystems will live on, after a fashion, as Libra Networks. In May of 2002, AMC bought a company called REBA Technologies, which had a patent pending for a technology to increase the speed of server operations. Using its experience working with high-end microprocessors and real-time analysis, as well as the cash it receives through the sale of its assets, what remains of AMC will pursue additional funding in order to sell and develop this technology under the Libra name.