PubMed1988
During the early 1980s, two major developments in computer technology changed the way chemists approached their science. The advent of the micropressor and then the PC changed experimental chemistry, while the availability of two classes of computer, the superminicomputer and supercomputer, greatly influenced computational chemistry. In the past two years, graphics workstation computers have begun to affect the practice of chemistry by combining fast, high-resolution, multiwindow graphics with superminicomputer power. In 1988, the advent of a new class of computer--the graphics supercomputer--offers extraordinary promise to both theoretician and experimentalist. In these systems, near-Cray compute power is combined with ultrahigh-speed 3-dimensional graphics for unparalleled visualization of molecular processes and other complex events. This is made practical not just by computing and graphics power but by use of ultrahigh internal bandwidths inside the graphics supercomputers. Another major development in scientific computing is the evolving concept of the laboratory computer network. Current network designs include hierarchical configurations incorporating various levels of computers--through supercomputers--either locally or via national or regional networks. New software methods are also having impact on chemical research, allowing, for example, the scientist to better abstract information from noisy or incomplete experimental data. Use of parallelism (multiple CPUs) in new design workstation computers will extend their power, by the early 1990s, past that of current supercomputer mainframes. Within five years the chemist will have $10 million of 1985 computer power on his desk, for considerably less than $100,000, along with visualization tools and software only dreamed of in 1985.(ABSTRACT TRUNCATED AT 250 WORDS)