Massively Parallel Technologies Hosts Discussion With Dr. Gene Amdahl

Gene Amdahl crafted "Amdahl's Law" in 1967, which states that in parallelization there are issues that will eventually place an upper limit on the maximal speed of the system, therefore mitigating much of the benefit of parallelization. MPT's breakthrough technology has solved these issues and Amdahl fully supports the company's breakthrough work in the field of supercomputing. During the presentation Amdahl spoke about the history of parallel processing, his revolutionary work in the field and endorsed MPT's work surrounding Amdahl's Law. "We really had something, and I was able to start a company that generated three or four billion in revenues. I see the same potential here. The sky's the limit," said Amdahl. Massively Parallel Technologies co-founder and chief technology officer Kevin Howard surprised Dr. Amdahl with the first mathematical derivation of "Amdahl's Law." Amdahl's Law, though never established from first principals, helped establish the supercomputing industry and has for more than 30 years been a force in the industry. "Amdahl's Law, and Dr. Amdahl himself, were already legends," stated Howard. "This mathematical derivation of Amdahl's Law cements his place in history. Dr. Amdahl's foresight is simply astonishing." Robert Duncan, Ph.D. also addressed the audience. Duncan is currently the Gordon and Betty Moore Distinguished Scholar within the Division of Physics, Mathematics, and Astronomy at the California Institute of Technology (Caltech). He is on leave from the University of New Mexico, where he is the Associate Dean for Research within the College of Arts and Sciences, and Professor of Physics. "The work that Kevin Howard has pioneered is some of the most exciting I've seen," stated Duncan. "The impact his work can have on the further growth of computer science is far more profound than the advent of personal computing." Massively Parallel Technologies' Virtual Power Centers mark the beginning of a new shift in computing where processing power is not contained in individual users' machines, but rather provided seamlessly through the Internet, on-demand.