Image compression multi-core synchronous fault-tolerance method, computer and processor

Abstract

The invention belongs to the technical field of hardware transient fault detection and recovery, and discloses an image compression multi-core synchronous fault-tolerance method, a computer and a processor. The method comprises the steps of constructing a synchronous image compression module, and making a single-core image processing program to conduct transplantation and optimization on a multi-core processor; according to existing core-level redundancy fault-tolerance characteristics, determining an unbounded asynchronous redundancy fault-tolerance model; according to the unbounded asynchronous redundancy fault-tolerance model, combining the synchronous image compression module to add a redundancy fault-tolerance function to an image compression system. According to the image compressionmulti-core synchronous fault-tolerance method, by achieving multi-core synchronous work on a main-core design management layer, the processor can be flexibly and conveniently transplanted on different kinds of hardware; the situations that there is a core being vacant to wait for data transmission and comparison is completed during synchronous redundancy fault tolerance are avoided; result comparison is conducted after two times of calculation; the extra expenditure caused by the data transmission is avoided, and memory access pressure of the system can also be reduced because the two operations are not overlapped in terms of time.

Description

technical field [0001] The invention belongs to the technical field of hardware transient fault detection and repair, and in particular relates to an image compression multi-core parallel fault-tolerant method, a computer and a processor. Background technique [0002] Space exploration activities require a lot of investment and high risk, and have extremely high requirements for computing reliability. The main factor affecting the safety of space probes in space is the radiation of cosmic rays, because there are a large number of high-energy particles including electrons, protons, particles and heavy ions in the cosmic environment. When the cosmic rays composed of these particles bombard the semiconductor of the aerospace computer When the circuit is damaged, it may cause a transient change in the stored power in the PN structure. This transient fault is also commonly referred to as the single event effect SEE (Single Event Effect). Although single event effects generally d...

AI Technical Summary

Problems solved by technology

In particular, the core-level redundancy fault tolerance currently has the following problems: (1) idleness in the redundancy causes performance waste: the core that completes the computing task first between the main core and the redundant core needs to wait for the other core to complete; (2) There is a waste of performance during the result verification: when the main core performs the result verification operation, the redundant core needs to be idle and wait for the result verification operation to complete, resulting in a waste of performance; (3) The binding mechanism between the main core and the redundant core is not flexible : Whether it is static binding or dynamic binding, the main core and redundant core work in pairs, which is not flexible enough, inconvenient to expand, and cannot adapt to the working state of an odd number of cores

[0003] To sum up, the problems existing in the existing technology are: the current nuclear-level redundancy fault tolerance has idleness in the redundancy, which causes performance waste; there is performance waste in the result verification; the binding mechanism between the main core and the redundant core is not flexible

The main reason for the problem is that the previously adopted multi-core parallel mechanism cannot fully meet the needs of fault tolerance. It is necessary to select an appropriate multi-merge parallel mechanism and design a multi-core parallel system that meets the needs of fault tolerance.

Images