Ball Semiconductor Inc has built its first spherical semiconductor, consisting of negative-channel metal-oxide semiconductor (NMOS) inverter circuits on a 1mm sphere. However, the pioneering firm is still finding it hard to get respect within the industry.
Allen, Texas-based Ball is developing a $5m pilot line to begin testing the mass production processes for spherical chips in mid-May. Initial production will concentrate on NMOS chips, but Ball hopes to move to production of the more popular and energy efficient CMOS semiconductors in June, with RF devices to follow. Currently the RF devices and 3D graphic accelorator markets are condisred the company’s chief targets, with standard CMOS logic device families also being important. Ball’s chief operating officer, Hideshi Nakano, says volume production will start late next year and forecasts that revenues will come on stream in 2001.
However, some industry watchers think that Ball’s spherical semiconductor plans are wildly over-optimistic. 4th Wave Analyst, John Latta, comments, Any technology that marks a major change in the infrastructure, i.e., fab, will have a difficult time. In an interview with our sister publication, Computer Business Review, MicroDesign Resources analyst, Jim Turley, said he considered the company an April Fool’s joke and claimed that its commercial rationale was unclear.
The cost and environmental benefits of the new process do seem attractive. Today’s wafer manufacturing processes waste up to 80% of the usable silicon, but Ball says spherical chips will see this cut to around 20%. A Ball chip fab would also replace today’s expensive clean rooms, by building devices within hermetically sealed pipes – a development that could reduce the construction cost of fabs by as much as 90%, the company claims.
But the company needs funding. It has so far raised $56m but needs roughly another $20m to build a full-scale production plant. Nakano refused to discuss future funding activities. In addition, Ball currently does not have the backing of any major US companies, something it needs if it is to have the pricing and marketing muscle to compete on the chip scene. It is apparent that one company cannot do everything or change the world, Nakano admitted, a consortium is essential and it is something that Ball is working towards.
In the longer term, Nakano has no clear view of where his company’s technology is going. It is difficult to foresee the future of Ball technologies, just as it was for anyone to foresee the impact of the internet or the invention of the first IC by Jack Kilby 40 years ago…Kilby himself, never dreamt that transistor cost would be cut into one-millionth, Nakano said. However, he is clear that those who say his company has bitten off more than it can chew are wrong. They [Ball’s critics] might be right in the sense that silicon is too solid to bite. They are wrong in the sense that Ball is as small as 1mm in diameter and easy to chew!
