In trying to speed up large-scale computations, the problem to solve is that of data movement: moving data around the system costs significant time and energy, and the problem very stubbornly refuses to go away (note that computer designers have been trying to solve the memory problem for over three decades). This talk will discuss several of the recent solutions that our group has helped to develop, including flash-based main memory systems and Micron's Hybrid Memory Cube DRAM. The talk will finish by looking at some of the remaining issues.
Bruce Jacob received the AB degree in mathematics from Harvard University in 1988 and the MS and PhD degrees in CSE from the University of Michigan in Ann Arbor in 1995 and 1997, respectively. He also worked for two successful startup companies: Boston Technology and Priority Call Management; at Priority Call Management he was the initial system architect and chief engineer. He is a Keystone Professor of electrical and computer engineering and former Director of Computer Engineering at the University of Maryland in College Park, and he is currently visiting at the University of Siena, Italy, where he is working on memory issues for manycore systems. He is a recipient of a US National Science Foundation CAREER award for his work on DRAM, and he is the lead author of an absurdly large tome on the topic of memory systems. His research interests include memory systems, operating systems, distributed systems, and designing electric guitars.
