Telcom Semiconductor

<p>Fundamental limits are still a ways off,” he said.</p>


PARIS — Recent regulatory changes have prompted research institutes and laboratories in Europe to spawn startup companies in unprecedented numbers, according to a Wednesday (March 6) panel at the Design Automation and Test in Europe (DATE) conference here.

Long considered a largely U.S. phenomena due to the ready availability of venture capital, that pattern of the past no longer applies, the panel said. Nearly half of electronics industry startups last year were founded in Europe, some said. The climate here has changed, and the worlds are coming together,” said Guido Arnout, chairman of CoWare Inc. (Santa Clara, Calif.).

CKCS5040-10uH/M_Datasheet PDF

The panel offered words of caution, as well. While European startups will likely find that European R&D engineers will work for less money and change companies less often than their U.S. counterparts, the transformation of researchers with academic roots into creators of successful products and businesses may be trickier and more time consuming than in America, they said.

You need to strike the right balance,” said Heinrich Meyr, chairman of Lisatek, who credited the stability of his company's R&D team in Aachen, Germany and the flexibility of its U.S. engineers in Menlo Park, Calif. Technology transferred from universities was a key to his startup's success, Meyr said. The universities provide us with people, research results and a test bed, while industry can give the universities focus, advice and funding.”

University roots

CKCS5040-10uH/M_Datasheet PDF

Many differences between U.S. and European startups grow from their different roots, with many of the former coming from industry, the latter from universities, panelists said. At universities, you are spending the money you already got from the government,” CoWare's Arnout said. In business, you spend money you haven't really earned yet — thus inviting VCs venture capitalists to breathe down your neck to come up with the results.” Moreover, European startups with university backgrounds must be prepared to make the big shift from developing technologies to developing solutions,” he said.

The real challenge” in Europe is that roughly 60 percent to 80 percent of its startups are founded by engineers from labs, said Eric Dupont, president and chief executive officer of Iroc Technologies (Santa Clara, Calif.). A close R&D relationship is the key to the future of your startup,” he told an audience crammed with engineers, young entrepreneurs and venture capitalists. But you need to be aware that the distance from technology to products is much longer for startups with laboratory roots.”

CKCS5040-10uH/M_Datasheet PDF

U.S. venture capitalists are less interested in pure technology than in commercializing technology, Meyr said. He advised startups to stay bootstrapped as long as possible — until you get the first paying customer or the first working silicon.”

Panelists also provided perspective on the location of a startup's headquarters. The location doesn't matter,” said Iroc's Dupont, except that you need to be closer to your customers. What you want is to become an industry way station.”

The Crolles2 facility and TSMC's Fab 12, in Hsinchu, Taiwan, will be the primary process-development sites for the new three-way partnership, with frequent exchanges of engineers between the locations, the companies said. TSMC has already validated SRAM test chips based on the 90-nm node that feature 1.3-mm2 cell sizes and a transistor channel length of 65 nm. Philips and ST also validated the 90-nm node technology in December by developing fully functional test chips — 1-Mbit and 4-Mbit embedded SRAMs with a density of 735 kbits/mm2 — produced on a pilot line using a 200-mm wafer.

In the near future, the partners intend to use the existing plants to validate the portable technology processes. Maintaining the independent 300-mm fabs in Hsinchu and Crolles is vital to assuaging customers' concerns about risk sharing and technological uncertainty, said Joel Monnier, corporate vice president and central R&D director at ST. We can offer fast access to volume capacity and compatible second sourcing from the beginning.”

Claasen said dual sourcing also should ramp up the yield rate faster. We will run exactly the same test chips on the two fabs, exchange the information between the fabs every six weeks and compare notes” on manufacturing tool quality, he said.

But F.C. Tseng, TSMC's deputy CEO, hinted that a 300-mm wafer fab joint venture may be a natural by-product of the relationship. For 300 mm, we don't rule out any possibility of cooperation in the future,” he said. When we move to 65 nm and we have an actual need for facilities, we will take [partnerships] into consideration.”

Thus far, about half of the companies doing 300-mm wafer development have gone solo or plan to. Aside from Intel and Samsung, they include Texas Instruments and DRAM makers Micron Technology, Nanya and Powerchip. Those pursuing joint projects or ventures have included Hitachi, Infineon and ProMOS Technologies. Motorola and foundry Chartered Semiconductor, meanwhile, have each delayed their 300-mm plans.

That's what Aura's chipsets and subsystems do: modulate a signal onto a coil, pull it off another coil and set it up for playback. They've patented a good deal of the clever signal processing that makes this trick possible at reasonable bitrates, as well as the design for an X-Y-Z antenna coil that works in any orientation to the transmitter.

Copyright © 苏ICP备11090050号-1 tl431 datasheet All Rights Reserved. 版权投诉及建议邮箱