NEC SX-8 Research Articles

Packaging technology for the NEC SX-3 Supercomputers : Hiroshi Murano and Toshikiko Watari. IEEE Transactions on Components, Hybrids, and Manufacturing Technology, 15(4), 411 (1992)

It is important for company to measure the consumer behaviour for the growth and increase market share. The COVID-19 crises have severe impacts on the purchase behaviour and consumption patterns of people, and their well-being, therefore the sociology of consumption has devoted attentions in the period of Crises. For this purpose, the data of 1700 customers on various dimensions were gathered using close ended-Likert scale questionnaire and analysed with Structured equation model through AMOS software. The outcome of the research is the Framework that include the predictors (i.e. Awareness of COVID-19, Attitude towards Pandemic and Product Type) and out comes (i.e. Consumer Behaviour, Change in Priority and Change in Method) due to threats that exhibited psychological process underlying reactance to their freedom. Results indicated the significant effects of Awareness of COVID-19, Attitude towards Pandemic, and type of product in shopping on changes in consumer behaviour and priorities in present crisis.

Performance of IBM RISC System/6000 workstation clusters in a quantum chemical application

The performance of clusters of IBM RISC System/6000 family of workstations in a quantum chemical application (electron correlation energy calculations using concurrent computation many-body perturbation theory) is assessed. The Parallel Virtual Machine System (PVM) is used to distribute various parts of the calculation among the workstations. The use of both homogeneous and heterogeneous clusters of workstations is investigated. The performance characteristics are compared with data recently obtained for the CRAY Y-MP C-90 and NEC SX-3/44 computers.

Parallel Processing and the Sustained Production Performance of the Cray Y-Mp: Benchmarks Using Optimized Microtasked Lattice Su(3) Code

The standardized-maximalist approach to supercom puter benchmarking consists in optimizing a standard production code on the supercomputer, then measur ing a wall-clock-based figure-of-merit that is relevant to users of the code in question. Since 1982, one highly efficient algorithm for simulating SU(3) lattice gauge theory has been used in such benchmarks, tracing the progress of supercomputers from the CDC 7600 to the CRAY X-MP and the NEC SX-2. Here we report on the performance of the CRAY Y-MP/8128 at the Cray Re search Corporate Computing Center under this bench marking procedure. The code was optimized and mi crotasked, taking advantage of the hardware and soft ware features of the Y-MP. The link-update time was measured with the code running on 1, 2, 4, and 8 CPUs. With 8 CPUs, it was 3.1 μsec. This corresponds to a sustained performance of 1.349 GFLOPS com puted on the basis of theoretical operation counts. (Hardware performance monitoring yields an estimate of 1.54 GFLOPS.) It represents an improvement of a factor 3.55 over a maximal CRAY X-MP configuration (four X-MP processors).

Processing of multi-dimensional NMR data with the new software PROSA

The new program PROSA is an efficient implementation of the common data-processing steps for multi-dimensional NMR spectra. PROSA performs linear prediction, digital filtering, Fourier transformation, automatic phase correction, and baseline correction. High efficiency is achieved by avoiding disk storage of intermediate data and by the absence of any graphics display, which enables calculation in the batch mode and facilitates porting PROSA on a variety of different computer systems; including supercomputers. Furthermore, all time-consuming routines are completely vectorized. The elimination of a graphics display was made possible by the use of a new, reliable automatic phase-correction routine. CPU times for complete processing of a typical heteronuclear three-dimensional NMR data set of a protein vary between less than 1 min on a NEC SX3 supercomputer and 40 min on a Sun-4 computer system.

Benchmarking High-Performance Computing Systems By Means of Local-Creutz Simulations of the d = 2 Ising Model

We have benchmarked various high-performance com puter systems by means of a code that solves the Ising model in two dimensions. We use a simple and porta ble Fortran implementation of the local-Creutz algo rithm. We compare the wall-clock times needed to per form 250,000 iterations of the algorithm on a lattice of 162 spins, for the Thinking Machines Corporation CM-2, the NEC SX-2, one CRAY Y-MP 4/32 processor, the Alliant FX/2800 with a variable number of proces sors, the IBM 3090-200 VF, an IBM 6000 RISC worksta tion, and a Silicon Graphics Iris workstation.

Photosensitive polyimide application for multi-chip packages.

This paper introduces the photosensitive polyimide application for the NEC SX-3 Supercomputer. In case of supercomputers, sophisticated packaging technologies are essential to achieve fastest operations as well as to use highest-speed, highly integrated LSIs.Wiring substrate which mounts and interconnects LSIs is the key to back up LSI's higher logical-operations.The high speed interconnection wirings and high density LSI mounting are requested for the substrate.The Multi-Chip Package using Photosensitive Polyimide had been developed to meet these demands.25μm wide, 75μm center-to-center spacings, four signal layers, 6ns/cm signal transmissions on a 225mm square substrate have been achieved by using this photosensitive polyimide multi layer substrate technology.

Packaging technology for the NEC SX-3 supercomputers

Since the performance of a computer system is mainly related to the machine cycle time, improvement of the logic circuit delay is the key for high-speed operations. In order to show how to reduce the logic circuit delay, the characteristics required for LSI chips and packaging technologies are discussed and the technological trends in packaging and their limitations are considered. As a typical application of high-performance packaging technology, the NEC SX-3 Supercomputer packaging is introduced, featuring 9 in/sup 2/ polyimide-ceramic substrates, a microchip carrier, flipped TAB carrier (FTC), high-density multichip packaging, high-speed coaxial cable interconnections, and a water cooling system.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Base- p-cyclic reduction for tridiagonal systems of equations

Cyclic reduction is often heralded as a vectorizable and fast method for the solution of weakly diagonally dominant tridiagonal systems of equations on a vector processor. However, in the standard version, where the number of equations is halved in each sweep (recursive doubling), sooner or later the algorithm runs into memory bank conflicts because the stride in the memory access is doubled in each sweep and eventually becomes a multiple of the number of memory banks. In this note we consider variants of the algorithm, in which the stride is tripled after each sweep and/or the reduced system of equations is moved to contoguous locations beyond the original ones. We shall compare performances on CRAY X-MP and NEC SX-2 machines.

A new Mersenne prime

The number 2 110503 − 1 {2^{110503}} - 1 is a Mersenne prime. There are exactly two Mersenne exponents between 100000 and 139268, and there are no Mersenne exponents between 216092 and 353620. Thus, the number 2 132049 − 1 {2^{132049}} - 1 has been verified as the 30th Mersenne prime in order of size.

Vector optimization of amber 3.0 on the NEC SX-2/400 supercomputer

Abstract The AMBER 3.0 molecular mechanics and molecular dynamics programs have been ported to and vectorized on the NEC SX-2/400 supercomputer. A detailed discussion of the vector enhancement of the AMBER non-bonded pair list generation subroutine is presented. Automatic vectorization using the FORT77SX compiler yielded speed-up factors of 1.2 to 1.5 over unvectorized code. Recoding of key portions of the program, as described in this paper, yielded speed-up factors of 1.8-2.7. The perturbation molecular dynamics program, PERDYN, now runs up to 35 times faster on the SX-2/400 than the VAX optimized version of the same program runs on the VAX 8650.

ARGOS, a vectorized general molecular dynamics program

AbstractThe organization of the highly vectorizable molecular dynamics simulation program ARGOS is described. The specific choice of the data structure and the separation of the calculation of interactions involving solutes and solvent molecules allows a considerable improvement in computation speed. Illustrative results are given for the NEC SX‐2/400 supercomputer. For the simulation of a large biological molecule in water a speedup factor of 5 is obtained as a result of vectorization of the code to 87%. The parts of the code used in a simulation of pure water could be vectorized to 98%, leading to an overall speedup factor due to vectorization of 13. The simulation of pure water runs over 300 times faster on the SX‐2/400 than on the VAX 8650.

Vector performance analysis of the NEC SX-2

This paper presents the results of a series of experiments to study the vector performance of the NEC SX-2. The main object of this study is to understand the architecture and identify its bottlenecks and limiting factors. A simple performance model is used to examine the impact of certain architectural features on the performance of a set of basic operations. The results of implementing this set on the machine for four vector lengths and three memory strides are presented and compared. These results show that the vector length and the ratio of floating point operations to memory references have a great impact on the performance of the machine. Two numerical algorithms are also employed and the results of these algorithms and the basic operations are compared to early results on one processor of the Cray-2 and Cray Y-MP. These comparisons show that the SX-2 is faster than the Cray Y-MP by up to 86% for short vectors and by 2 to 4 times for long vectors. Also, it outperformed the Cray-2 by even bigger factors. Finally, the architecture of the SX-X is presented, and some predictions about its performance are given.

Three-Dimensional Dip Moveout On the Sx-2: an XMU Implementation

In the past few years, a seismic processing technique called dip moveout (DMO) has been investigated with the objective of improving the quality of seismic data. The main memory requirement of a three-dimensional DMO program with a realistic data set may be on the order of several gigabytes, a requirement that none of today's supercomputers can satisfy. One solution may be to implement out-of-core algorithms, which process only a subset of data at any time. So, only (part of) the instructions and part of the data need to reside in the main memory. The entire program data could reside on a secondary storage unit. The NEC SX-2 supercomputer used in this study has a main memory size of 128 me gabytes. It also has an extended memory unit (XMU) size of 1 gigabyte. The XMU is connected to the main memory unit by a 1.3 gigabyte/sec channel. An out-of- core algorithm has been implemented on the SX-2. The XMU was used to store the program data. The program was tested with three-dimensional data from an acoustic tank used for physical modeling. The results indicate that the speedup between the SX-2 and the VAX-11/780 is in excess of 100.

Numerical Analysis on Two-Dimensional Flow and Heat Transfer of Louvered Fins Using Overlaid Grids

Using an overlaid grids method, finite difference analysis have been made on two-dimensional flow and heat transfer characteristics of louvered fins which comprise automotive heat exchangers used in the forced convection regime at low Reynolds number. In this method the physical domain is subdivided into some regions which can adapt easily to Cartesian grids. The maximum total number of the grid points allocated non-uniformly in the respective regions is about 100000. Communication between the grids is accomplished by linear interpolation of the dependent variables at grid boundaries. It took 3060 minutes of the CPU (supercomputer NEC SX-2) for the iterative computation. In the rage of Re=64450, the present numerical scheme was validated through the comparisons between the numerical results and the corresponding experimental data which were also obtained in the present study.

Optimizing the Su(3) Lattice Gauge Theory Algorithm On the Nec Sx-2 Supercomputer

The SU(3) lattice gauge theory algorithm based on the Metropolis method is vectorized and optimized on the NEC SX-2 supercomputer, a vector uniprocessor, and a link-update time of 14.8 μsec is obtained. In the version of the Metropolis method with the checkerboard data ordering used on a Fujitsu VP-200, the number of calculations needed to evaluate the action from each plaquette was 76 multiplications and 24 additions, while in the new version for the NEC SX-2, the number of cal culations is 18 multiplications and 24 additions. The NEC SX-2 algorithm is 1.2 times faster than the algo rithm on the Fujitsu VP-400 supercomputer system, 2.7 times faster than on a single processor of the CRAY X-MP, and 1.8 times faster than on a four-pipeline CDC Cyber 205.

Factoring with the quadratic sieve on large vector computers

The results are presented of experiments with the multiple polynomial version of the quadratic sieve factorization method on a CYBER 205 and on a NEC SX-2 vector computer. Various numbers in the 50–92 decimal digits range have been factorized, as a contribution to (i) the Cunningham project, (ii) Brent's table of factors of Mersenne numbers, and (iii) a proof by Brent and G. Cohen of the non-existence of odd perfect numbers below 10 200. The factorized 92-decimal digits number is a record for general purpose factorization methods.

Numerical analysis on two-dimensional flow and heat transfer of louvered fins using overlaid grids.

Using an overlaid grids method, finite difference analysis have been made on two-dimensional flow and heat transfer characteristics of louvered fins which comprise automotive heat exchangers used in the forced convection regime at low Reynolds number. In this method the physical domain is subdivided into some regions which can adapt easily to Cartesian grids. The maximum total number of the grid points allocated non-uniformly in the respective regions is about 100000. Communication between the grids is accomplished by linear interpolation of the dependent variables at grid boundaries. It took 3060 minutes of the CPU (supercomputer NEC SX-2) for the iterative computation. In the rage of Re=64450, the present numerical scheme was validated through the comparisons between the numerical results and the corresponding experimental data which were also obtained in the present study.

Polyimide-Ceramic Substrate for Supercomputer Packaging

AbstractThis paper introduces the Polylmide-Ceramic substrate for NEC SX Supercomputers. In case of high performance system such as supercomputers and top end machines in general purpose computer, sophisticated packaging technologies are essential to achieve fastest operations as well as to use highestspeed, highly integrated LSIs.Wiring substrate which mounts and interconnects LSIs is the key to back up LSI's higher logical-operations.The high speed interconnection wirings and high density LSI mounting are requested for substrate.The Polyimide-Ceramic substrate had been developed to meet these demands and have many features of high density thin film wiring, high power supply, high thermal conductivity and huge number of I/Os, in addition to high speed wiring.25μm wide 75μm center-to-center spacing, two signal layers, 6ns/m signal transmissions, 2.5W/cm2 high power density, 2177 I/Os on a 100mm square substrate have been achieved by using this super substrate technology.The packaging hierarchy, the first level packaging of TAB LSI, the second level of multi-chip packaging by using Polyimide-Ceramic substrate and liquid cooling module, and the third level of board assemblies are introduced.

Implementation and Performance of the Large-Scale Finite Element Code Sesam On a Wide Range of Scientific Computers

SESAM is a large, commercial finite ele ment package for structural analysis that includes SESTRA, an analysis portion of approximately 280,000 lines of FORTRAN 77 statements. We discuss the issues of por tability and efficiency in implementing this code on a wide range of computers, in cluding the FPS-164, GRAY X-MP, CDC Cyber 205, Amdahl VP1200, IBM 3090/VF, and NEC SX-2. Performance data are re lated to the different architectures and the degree of code tuning. SESAM requires large amounts of direct access secondary storage. The use of FORTRAN I/O can sig nificantly decrease performance, as mea sured by both CPU time and elapsed time. We discuss how different levels of I/O support on scientific computers affect the potentially large gap between CPU time and elapsed time.

Packaging technology for the NEC SX supercomputer : Toshihiko Watari and Hiroshi Murano. IEEE Trans. Compon. Hybrids mfg Technol.CHMT-8 (4), 462 (1985)

Packaging technology for the NEC SX supercomputer : Toshihiko Watari and Hiroshi Murano. IEEE Trans. Compon. Hybrids mfg Technol.CHMT-8 (4), 462 (1985)