About ICSI Research Groups Projects Publications Events Partnerships Visitor Programs News Search
Algorithms AI Networking Speech Architecture Vision Other Activities
Members    
 

Research Groups

Architecture

 
 

Professor Krste Asanovic moved from MIT to join the EECS faculty at UC Berkeley in Summer 2007, and has started a new Architecture Group at ICSI. The major focus of the group is the realization of efficient parallel programmable architectures exploiting advances in circuit and device technologies. The group works closely with the ParLab at UC Berkeley, developing a new parallel software stack, and the RAMP consortium, developing FPGA-based parallel architecture emulators.

Current Projects:

Monolithic Silicon Photonics for Processor-to-DRAM Interconnects

In a collaboration with the MIT Center for Integrated Photonic Systems, researchers from the Architecture Group are exploring the use of silicon photonics for processor-to-memory interconnect. Projected advances in electrical signaling seem unlikely to fulfill the memory bandwidth demands of future manycore processor chips. Monolithic silicon photonics, which integrates optical components with electrical transistors in a conventional CMOS process, is a promising new technology that could provide large improvements in achievable interconnect bandwidth. In a DARPA-funded effort, the ICSI Architecture Group is exploring possible memory interconnect schemes to exploit the new photonic device and circuit technology being developed at MIT.

Maven (Malleable Array of Vector-thread ENgines)

In earlier work at MIT, Professor Asanovic's team developed the Scale vector-thread architecture and processor prototype, which combines data-level and thread-level parallel execution models in a single unified architecture. Maven is the second-generation vector-thread architecture, designed to scale up to hundreds of execution lanes, and with the goal of providing very high throughput at low energy for a wide variety of parallel applications. Maven is based on a new compact lane design, which is replicated to yield a "sea-of-lanes" execution substrate. At run-time, lanes are ganged together to form variable-sized vector-thread engines, sized to match application needs.

Browse Architecture Group Publications

top

Copyright © 2008 International Computer Science Institute. All Rights Reserved.
Untitled Document