The Armstrong III Reconfigurable Computing Project

Armstrong III is 20-node parallel computer with Custom Computing Capabilities. Armstrong III was built to study the feasibility of augmenting the computational capabilities of a RISC microprocessor by adding an FPGA-based reconfigurable coprocessor. Armstrong III is currently being used as a computational platorm to accerelate Hidden-Markov speech-model training. Armstrong III has also been used for

• Genetics Algorithms for Chip Partioning
• Discrete Cosine Transform Engine
• Image Processing

The hardware and operating system for Armstrong III were designed and built in LEMS . Here are a few pictures of Armstrong III.

Publications

• Processor Reconfiguration Through Instruction Set Metamorphosis: Compiler And Architecture
• PRISM-II Compiler and Architecture
• An Asynchronous Approach to Efficient Execution of Programs on Adaptive Architectures Utilizing FPGAs
• An Asynchronous Approach to Synthesizing Custom Architectures for Efficient Execution of Programs on FPGAs
• Implementing a Genetic Algorithm on a Parallel Custom Computing Machine
• Performing Log-Scale Addition on a Distributed Memory MIMD Multicomputer with Reconfigurable Computing Capabilities

Tech Reports

• Armstrong III: A Loosley-Coupled Parallel Processor with Reconfigurable Computing Capabilities

Back to the LEMS homepage