Computer Architecture and Automated Design Lab

The CAAD lab is a research group in Boston University's Department of Electrical and Computer Engineering, and is under the direction of Prof. Martin Herbordt.

Hot News

(news archives)

March, 2009: Martin Herbordt gave the keynote talk at Many-Core and Reconfigurable Supercomputing Conference in Berlin, Germany.

February, 2009: Martin Herbordt and Miriam Leeser are co-chairing the ASAP 2009 conference, being held in Boston.

Research Focus

Our research begins with computer architecture, which we like to define (very broadly) as the application of technology to solve problems in data processing. Our particular focus is on applications that are likely to benefit from hardware solutions other than commodity systems. Our current primary interest is in the use of FPGAs as computational coprocessors to solve problems in bioinformatics and computational biology (BCB). There are two aspects to this work: demonstrating the benefits of FPGA computation with respect to particular problems in BCB and creating a software environment to allow transparent application development.

As with most projects in experimental computer systems, ours requires integration of disparate disciplines within computer science and engineering. For us, these include computer architecture, design automation, circuit design, HL and HD languages and compilers, mapping and portability, algorithms, and application analysis.

Facilities

We are located in the Photonics Center Building at 8 Saint Mary's Street. In our research labs we maintain:

• A number of workstation class PCs as well as Sun workstations used primarily for EDA computations
• EDA software packages from Synopsys, Cadence, Synplicity, ModelSim, Xilinx and Altera
• Various FPGA development boards including Wildstar II Pro™ from Annapolis Microsystems, featuring dual Xilinx XC2VP70s Virtex-II Pro™ chips
• A four-processor Altix 350 with RASC reconfigurable accelerator from Silicon Graphics
• A dual-Opteron XD1000 development system with XD1000 FPGA coprocessor module from XtremeData

Support

The lab has recently benefitted from the support of:

• IBM faculty research award
• National Institutes of Health (NIH) grant: FPGA-Based Computational Accelerators
• U. S. Naval Research Laboratory
• MIT Lincoln Laboratory
• Synplicity, ModelSim, Synopsys and Cadence, through their university programs.
• Xilinx
• Silicon Graphics
• XtremeData