American universities need greater access to supercomputers
Gene Dallaire · Communications of the ACM · 1984
The vast majority of researchers at American universities have no access to supercomputers.What are the consequences of that fact on the quality of American research and on the future dynamism of American industry?What specific steps should be taken to correct the problem? GENE DALLAIREFor the past three summers, Larry Smarr, a professor of astrophysics at the University of Illinois, has packed his bags and journeyed--together with a handful of his graduate students--to the Max Planck Institute for Physics and Astrophysics in Munich, West Germany.There, Smarr plunges into his simulations of large radio emission jets, which stream outward from certain galaxies, on the Institute's American-made Cray-1 supercomputer.He gets all the time he needs on the supercomputer--at no charge.The Institute, funded jointly by the German Government and German industry, feels adequately compensated by the opportunity to be a partner in significant research work.Smarr is not the only American university researcher traveling to Europe these days to get access to a supercomputer.German universities have become quite a mecca for American university researchers in astronomy and other fields.The Smarr case, then, is not an isolated one.And it points to a problem that is causing increasing anxiety in U.S. university, industry, and government circles--the fact that the overwhelming majority of researchers at American universities have no access to what is increasingly being recognized as one of the most important research, development, and design tools of the late 20th century: the supercomputer.