Hypercube computer

The first phase of Hypercube s existence involved contracting for Intel in association with Intel s development of the first commercial highly parallel computer, the Intel iPSC hypercube computer. This is the origin of the com-... 

New machine architectures are performing remarkably, provided the software can be written to use the architecture. Hypercube computers available commercially are multidimensional arrays of hundreds of computers, each with the power of a microvax. 

Smith has also described the implementation of MD on parallel machines, with particular reference to hypercube computers. In work designed to introduce a novice to the main aspects of parallel computing in MD, Smith described three particular algorithms replicated data (RD), systolic loop (SLS-G), and parallelized link cells (PLC), all of which have good load balancing. The performance characteristics of each algorithm and the factors affecting... 

S. E. DeBolt and P. A. Kollman, /. Comput. Chem., 14, 312 (1993). AMBERCUBE MD, Parallelization of AMBER S Molecular Dynamics Module for Distributed-Memory Hypercube Computers. 

Because of its simplicity, efficiency and adaptability to the hypercube computer architecture being developed at the California Institute of Technology,we have chosen Johnson s logarithmic derivative method to numerically integrate eq. (6.3). 

Smith W 1991 Molecular dynamics on hypercube parallel computers Comput. Phys. Commun. 62 229-48... 

W. Smith, Molecular dynamics on hypercube parallel computers , Comp. Phys. Comm., Vol 62, no 2 3, 229-48, 1991. 

The HyperChem philosophy associated with back end computations is one which is in tended to in still eon fiden ce. as far as is possible. in the scientific results emanating from HyperChem. This ph ilosoph y is on e of open n ess — open n ess aboii t the prod net. the calculations being performed, the science embodied in the product, etc. Apart from protecting the proprietary code associated with a commercial product. Hypercube wushes to document and describe as fully as is possible tb e calculation s th at HypcrCb cm performs. There should be no mystery about the scientific results obtained with HyperChem. 

Iman, R. L. and M. J. Shortencarrier, 1984, A FORTRAN 77 Program and User s Guide for the Generation of Latin Hypercube and Random Samples for Use with Computer Model, NUREG/CR-3624, March. 

Hypercubes and other new computer architectures (e.g., systems based on simulations of neural networks) represent exciting new tools for chemical engineers. A wide variety of applications central to the concerns of chemical engineers (e.g., fluid dynamics and heat flow) have already been converted to run on these architectures. The new computer designs promise to move the field of chemical engineering substantially away from its dependence on simplified models toward computer simulations and calculations that more closely represent the incredible complexity of the real world. 

Hypercube, Inc. at http //www.hyper.com offers molecular modeling packages under the HyperChem name. HyperChem s newest version, Hyper-Chem Release 7.5, is a full 32-bit application, developed for the Windows 95, 98, NT, ME, 2000, and XP operating systems. Density Functional Theory (DFT) has been added as a basic computational engine to complement Molecular Mechanics, Semiempirical Quantum Mechanics and ab initio Quantum Mechanics. The DFT engine includes four combination or hybrid functions, such as the popular B3-LYP or Becke-97 methods. The Bio+ force field in HyperChem represents a version of the Chemistry at HARvard using Molecular Mechanics (CHARMM) force field. Release 7.5 of HyperChem updates... 

We can think of these representations as shadows of hypercubes on 2-D pieces of paper. Luckily, we don t have to build the object to compute what its shadow would look like. (The computer code I used to create these forms is listed in Appendix I.) Projections of higher-dimensional worlds have stimulated many traditional artists to produce geometrical representations with startling symmetries and complexities (Figs. 4.18 to 4.20). 

Lastly, we would like to mention here results of the two kinds of large-scale computer simulations of diffusion-controlled bimolecular reactions [33, 48], In the former paper [48] reactions were simulated using random walks on a d-dimensional (1 to 4) hypercubic lattice with the imposed periodic boundary conditions. In the particular case of the A + B - 0 reaction, D = Dq and nA(0) = nB(0), the critical exponents 0.26 0.01 0.50 0.02 and 0.89 0.02 were obtained for d = 1 to 3 respectively. The theoretical value of a = 0.75 expected for d = 3 was not achieved due to cluster size effects. The result for d = 4, a = 1.02 0.02, confirms that this is a marginal dimension. However, in the case of the A + B — B reaction with DB = 0, the asymptotic longtime behaviour, equation (2.1.106), was not achieved at all - even at very long reaction times of 105 Monte Carlo steps, which were sufficient for all other kinds of bimolecular reactions simulated. It was concluded that in practice this theoretically derived asymptotics is hardly accessible. 

Hypercube s HyperChem is the best-known computational chemistry software package developed in Canada and marketed worldwide. Hypercube, Inc., the brainchild of Neil Ostlund, was incorporated in February 1985. The company has a headquarters in Waterloo, Ontario, adjacent to the University of Waterloo, but moved most of its operations to Gainesville, Florida, in 1997. 

Hyperchem, Inc., Hyperchem Users Manual Computational Chemistry, Hypercube, On, Canada, 1994. 

Iman RL, Shortencarier MJ (1984) A FORTRAN 77 program and user s guide for the generation of Latin hypercube and random samples for use with computer models. Albuquerque, NM, Sandia National Laboratories (Report Nos. SAND83-2365 and NUREG/CR-3624). 

Massively parallel (multiple instruction, multiple data) computers with tens or hundreds of processors are not readily accessible to the majority of quantum chemists at the present time. However the cost of currently available hypercube machines with tens of processors (each with about the power of a VAX) is comparable to that of superminis but with up to a hundred times the power. For applications of the type discussed above the performance of a machine with as few as 32 or 64 processors would be comparable to (or perhaps even exceed) that of a single processor supercomputer. Although computer requirements currently limit QMC applications (even with effective potentials) the proliferation of inexpensive massively parallel machines could conceivably make the application of relativistic effective potentials with C C quite competitive with more conventional electronic structure techniques. 

The four-index transformation is a good test case for parallel algorithm development of electronic structure calculations, because it has O(N ) operations, a low computation to data transfer ratio and is a compact piece of code. Distributed-memory algorithms were presented for a number of standard QC methods by Whiteside and co-workers Li52 special emphasis on the integral transformation. Details of their implementation on a 32-processor Intel hypercube were provided. 

---