|
|||||
Sweden SoC effort finds Taiwanese ally
Sweden SoC effort finds Taiwanese ally LONDON A Taiwanese company is poised to join Sweden's nationally supported system-on-chip R&D program, known as SOCware. The unidentified company will work with Acreo AB (Norrkoping), a Swedish national institute, to establish a design group specializing in broadband wireless communications within the Acreo research institute. The formation of a group of about 15 engineers is planned, following a model currently being worked through with Atmel Corp. (San Jose, Calif.), which announced its involvement in SOCware last June. The Taiwanese company will gradually take over the design group. Rolf Rising, head of the SOCware program at the Invest in Sweden Agency (Stockholm), declined to identify the company but said that access to Taiwanese deep-submicron CMOS manufacturing will be a useful additional resource for the SOCware initiative in its attempt to foster a cluster of system-level design companies in Sweden. In the 18 months since the initiative was formally launched, it has announced only one outside investor. But Atmel's inclusion in SOCware is claimed to be strategic, since the company can provide access to silicon-germanium chip manufacturing to the universities, research groups and, potentially, fabless design groups that join SOCware. Despite the paucity of investors, Rising said the program is proceeding with a long-term philosophy of investing in education, research and strategic partnerships to stimulate investment. To that end, Linkoping and Lund Universities and the Royal Institute of Technology in Stockholm are due to qualify 50 master's graduates in system-level integration engineering in 2001, 100 in 2002 and 150 in 2003. Under the SOCware partnership model, Acreo builds up a design group that works for an investor. At the end of two years, group members become employees of the foreign investor. The investor, if it becomes dissatisfied with the arrangement, has the option of pulling out and taking the intellect ual property with it. At its launch, SOCware's goal was to create some 1,500 engineering jobs within four or five years. Thus far, Atmel has announced its intention to form a design group of about 20 engineers. "Job count is not what is driving us," said Rising, although he added that the original goal of creating 1,500 jobs could still be attained within two and a half to three and a half years. "We can offer design teams, house those teams during the startup phase and even provide some cost-sharing of the R&D." Up to 40 percent of the cost can be borne by the Swedish government, he explained, since the SOCware cluster is a public project. The government also has pledged to spend $60 million on the project over about five years. According to Acreo sales manager Marten Armgarth, one reason for the slight slowdown in system-level work at Acreo in the past two years has been the establishment of Bluetronics AB (Norrkoping). Formed in 1999, Bluetronics now employs 40 people, most recruited from Acreo. That has forced Acreo to rebuild but serves as evidence that the agency's job-creation strategy is working, said Armgarth. SOCware, launched in September 1999, is somewhat akin to Project Alba, a similar program based in Scotland and launched in December 1997. Alba started with a commitment from Cadence Design Systems Inc. (San Jose, Calif.) to create nearly 2,000 design jobs, which later transferred to Tality Inc. as Cadence design services were spun off into that organization. The intention was that, with Tality as a seed company, smaller design groups and spin-offs would be created, with a focus on system-level integration. While education, through the Institute for System-Level Integration (ISLI), is a key part of Project Alba, the academic unit started on a smaller scale than SOCware's, with eight master's students in 1999 and a doubling of master's program participants in 2000. Another 16 students have started a four-year engineering degree at ISLI. Both the Scottish and Swedish appr oaches have shown modest progress for a clustering process that is generally acknowledged to require many years to succeed.
|
Home | Feedback | Register | Site Map |
All material on this site Copyright © 2017 Design And Reuse S.A. All rights reserved. |