Alvin R. Lebeck
Professor of Computer Science

Computer Systems Engineering with a focus on computer architecture, nano-scale systems, memory system, energy efficient computing, and multiprocessors.

Current Research Interests

Hello web surfers, welcome to my corner of the internet.  I carry out research and teach in the broad area of computer systems, especially in computer architecture. My research spans from atoms to applications. Below is a bit more about my interests.

Accelerating Statistical Machine Learning
Novel computer architectures targeted at machine learning problems that rely on Markov Chain Monte Carlo techniques and probabilistic computing in general.

Multicore/multiprocessor and Distributed Systems (including data centers)
Cache coherence protocols, memory consistency, runtime systems to expose and exploit parallelism. I also explore the use of new hardware technology in data centers.

Memory Systems 
General aspects of memory hierarchies such as cache design, prefetching, latency tolerance, etc.

Architectures for Emerging Technologies  
DNA self-assembled systems, self-organizing systems, defect-tolerance, nanophotonics.  Although my active research in this area is winding down, I do have a company (Phitonex) that is commercializing some of the technology in the life science space.  I continue to be interested in new computational paradigms enabled by new technologies.

Energy Efficient Computing 
Processor and memory design, operating system support.

Check out Computer Architecture Today the ACM SIGARCH Blog (I was the founding editor).

If you are in the life sciences you may interested in Phitonex, a company based on our years of research on DNA nanostructures and fluorescence.

Current Appointments & Affiliations

Contact Information

Some information on this profile has been compiled automatically from Duke databases and external sources. (Our About page explains how this works.) If you see a problem with the information, please write to Scholars@Duke and let us know. We will reply promptly.