Particle model faced multi-tier direct-connection cluster paralleling computing system

Abstract

The invention relates to the field of high-performance computer value simulation technology and discloses a particle model-oriented multi-layer direct-connecting cluster parallel computer system. The system comprises a plurality of nodes; the nodes are arranged into one-dimensional or multi-dimensional array of a layer or multiple layers in logic; communication connection can be directly carried out between the adjacent nodes on the same layer, thus realizing the high-speed data transmission between the adjacent nodes on the same layer and the communication connection between nodes on different layers through an exchanger. The invention simplifies the design of the high-performance computer system, increases the calculation rate of the high-performance computer system on the particle method and similar calculations, and avoids the conflict between the communication and the memory access, thus fully overlapping the calculation and the communication operation of the nodes, allowing the calculation to gradually advance towards the sequential time from the peripheral area corresponding to the nodes to the center and further effectively relieving the small-scaled load fluctuation.

Description

technical field [0001] The invention relates to the technical field of high-performance computer numerical simulation, in particular to a particle model-oriented multi-layer directly connected cluster parallel computing system. Background technique [0002] At present, the mainstream way to achieve high-performance computing is massively parallel processing (MPP). Since such a system costs a lot, how to make full use of its hardware resources is an important consideration in the design of MPP. The traditional way of thinking is to make the system adapt to many different algorithms and application problems, that is, to rely on versatility to ensure the fullness of business volume and make full use of system resources. [0003] For this reason, global fast data exchange should be realized in principle, including direct or indirect data exchange between processors and memories and between processors. Under such a design idea, when the number of processors increases, the hardwa...

AI Technical Summary

Problems solved by technology

At present, even for single-processor microcomputers, the widening speed gap between the data processing speed of the processor and the data access speed of the memory makes it only about 10% of the peak value of the actual efficiency of running many calculation programs. For MPP systems, it is often reduced to only 1-2%

On the other hand, although there are some special-purpose parallel computers for specific algorithms and problems, they cannot be mass-produced and fully utilized because of their narrow application and small business volume, and the economy is very poor, so they have not been focused on development.

Images