Clint Schow Seminar: Moving Photonics Beyond Wire Replacement
Manager, Optical Link and System Design Group
Moving Photonics Beyond Wire Replacement
March 7, 2014 | 1:00pm | ESB 1001
Faculty host: Larry Coldren
The big opportunity for photonics is now Computercom: developing the interconnection networks that will underpin the next generations of high performance computers (HPC) and data centers. The requirements for optical hardware demanded by these applications are almost completely opposite from the conventional Telecom market where performance and capacity are achieved at nearly any cost, power or complexity. In contrast, as there are current HPC machines with > 500,000 optical links, Computercom research must focus on developing technologies that can enable millions of high-speed links in future systems that simultaneously offer low cost, low-power consumption, and high bandwidth density. At IBM Research, we have been actively addressing this challenge by building hardware targeted at integrating optics deeply into systems. Some examples include Tb/s parallel optical transceivers, optical printed circuit boards incorporating polymer waveguides for chip-to-chip links, and monolithically integrated Si photonics for transceivers and switches. We have pushed the limits of speed and power efficiency, demonstrating the fastest VCSEL link (64 Gb/s) and best wall-plug power efficiency for any high-speed optical link in any technology (1 pJ/bit at 25 Gb/s). We continue to work on improving the performance of Computercom components and expect to displace copper at ever-shorter distances in future systems. But as long as optics only offers a better “wire,” the choice between optical and electrical interconnects largely hinges on challenging cost and power consumption comparisons. The next wave of photonics research should therefore focus on exploiting unique features of photonic technologies to gain application-level advantages, opening new opportunities for the proliferation of optics beyond the traditional and ongoing replacement of copper.
Clint L. Schow received B.S., M.S. and Ph.D. degrees in electrical engineering from the University of Texas at Austin. In 1999, Dr. Schow joined IBM in Rochester, MN, assuming responsibility for the receivers used in IBM’s optical transceiver business. From 2001 to 2004, he was with Agility Communications in Santa Barbara, CA, developing high-speed optoelectronic modulators and tunable laser sources. In 2004, Dr. Schow joined the IBM T.J. Watson Research Center as a Research Staff Member where he conceived and designed generations of VCSEL-based links and parallel optical transceivers that set records for power efficiency, data rate, and bandwidth density. He was a primary contributor to IBM’s monolithic Si photonic/CMOS co-design efforts and led the circuit designs for the first generation of transceivers and photonic switch chips.
Dr. Schow is currently the Manager of the Optical Link and System Design group responsible for optics in future generations of high performance computers and datacenters. He has led numerous cross-department, multi-site joint R&D projects involving extensive internal, external, and international collaborations. As PI for multiple DARPA-sponsored programs, Dr. Schow has directed investigations into chip-to-chip optical links, nanophotonic switches, and new system architectures based on low-latency photonic switching fabrics. He is a Senior Member of the OSA and the IEEE, has published more than 140 journal and conference articles, and has fourteen issued and more than fifteen pending patents.