San Francisco. Imagine that you could assemble a team consisting of a brilliant physicist, an entrepreneur with character, a genuine marketer, a pioneer, a go-getter, a chemist, a businessman, a trendsetter, and a steady hand.
If you could do that, the unexpected would happen, said Luc Van den hove, imec president and CEO, as he presented images of people including Albert Einstein, Steve Jobs, Margaret Thatcher. Van den hove made the remarks in his opening address July 8 at the 2013 Imec Technology Forum US, held in conjunction with SEMICON West.
He noted that managing such a team may be difficult, but it would be worth it, because when you bring together people with complementary skills, innovation occurs at the crossing of disciplines.
“That's what we try to do at imec,” Van den hove said. “We try to create a 'sum of minds' and shift the boundaries of technologies.”
Van den hove provided examples related to three technologies: a cell sorter, logic and memory scaling beyond 10 nm, and plastic displays.
The cell sorter looks for one cancer cell among a billion normal cells. Today, cell sorting requires a manual approach using bulky, expensive equipment. Imec, in contrast, applies engineering knowhow to develop a high-content, high-throughput cell sorter the size of a chip. The sorter employs a microfluidic channel the width of a cell, and thanks to silicon technology, Van den hove said, imec can deploy 1,000 channels in parallel. The imec sell sorter brings analysis “from bench to bed,” he said.
The cell sorter involves many disciplines, including CMOS processing, wafer-level microfluidics, cellomics (the discipline of cell biology and behavior), high-speed CMOS imager design, wave optics modeling, ASIC design, and MEMS fabrication expertise to implement microfluidic bubble switches. The cell sorter, he said, bridges the gap between nanoelectronics and life sciences, revolutionizing life-science research, diagnose illnesses, and treat patients. (EE-Evaluation Engineering's August issue will feature a related article on imec's work in the life-sciences area.)
With respect to logic and memory scaling, Van den hove discussed a variety of devices and architectures, including the Ge FinFET, the III/V FET, the gate-all-around FET, the vertical FET, the tunnel FET, and graphene nanowires. The road to 10-, 7-, and 5-nm geometries, he said, will require combined expertise in physics, materials, device architectures, and process integration. For decades, he said, lithography has driven geometry scaling, but that's no longer enough. “We need new materials and device structures,” he said, such as the high-mobility FinFET.
He also described work involving an ultrathin hybrid floating-gate cell that could extend flash memory to 15 nm, and 3D SONOS could extend flash down to 10 nm. In addition, DRAM will scale beyond 20 nm, with devices such as the STT MRAM. He described various heterogeneous devices, with stacked DRAM and logic; analog and digital; DRAM plus logic, ASPs, analog, and I/O; and analog plus MEMS sensors stacked on digital logic. 3D implementations, he said, require expertise in applications, design, and process technology. We have reached the end of easy scaling, he said.
Van den hove also discussed patterning, contrasting double patterning with today's lithography with single patterning using extreme ultraviolet, or EUV. Despite delays, he said, EUV will be needed to continue scaling while reducing cost per gate with 450-mm wafers.
Van den hove wrapped up his presentation by describing flexible plastic displays, which require expertise in disciplines including flexible and stretchable electronics, low-power design, material process integration, and characterization and modeling. Flexible displays will be built on plastic foils instead of glass substrates, he said, and will require processing at lower temperatures. Sustained R&D will enable the plastic electronics revolution, he said.
Imec, he concluded, brings together people of 75 nationalities to make the unexpected happen.