University of Southern California Center for Interactive Smart Oilfield Technologies The USC Andrew and Erna Viterbi School of Engineering USC
John Granacki




























.

John Granacki is the Director of the Advanced Systems Division (ASD) at USC/Information Sciences Institute. He has over 33 years of industrial and research experience including work in the development of computer systems, integrated circuits, and design tools. He received his Ph.D. in Electrical Engineering in 1986 at the University of Southern California. He worked at Hughes Aircraft Company for 10 years during which time he developed large computer systems, special purpose processors, VLSI chips and CAD tools for VLSI and PCB layout. He worked for 7 years at RCA Solid State Division where he was a member of the LSI microprocessor development department. He designed a variety of LSI/VLSI circuits including custom I/O devices, A/D converters and memories.

Under Dr. Granacki’s leadership and technical direction, ASD has developed a wide spectrum of innovative systems. These systems include DIVA, a new computer architecture based on PIM technology Package-Driven Scalable System, an embeddable computer using MCM technology and 3-D packaging to demonstrate a compute density of 100 GFlops per cubic foot and the GUMPS Brick, a hand-held wireless computing node with modular capability.

Dr. Granacki’s current research is focused on the architecture of embedded systems for computing and communication. His research group are working on creating hybrid architectures that can perform efficiently under different computing paradigms. For example, on the DARPA funded MONARCH Project, they have developed two new concepts: Data Flow In Memory (DFIM) and SPIM (Standalone Processor In Memory) that can be implemented as a single machine that runs efficiently for both streaming applications and conventional applications that run under threaded control. He is also exploring biologically-inspired architectures including advanced Dynamic Synapse Neural Networks with Dr. Ted Berger in Biomedical Engineering. They are researching efficient VLSI implementations for a wide-range of applications involving temporal-spatial pattern-recognition. He is also interested in Design Automation, reaserch, specifically at the system-level.



Students:

Chiu Hsien Chan Thumbnail
Chiu-Shien "Jerry" Chan
Xiang Fang Thumbnail
Xiang "Frank" Fang