Gene data interpretation-oriented reconfigurable heterogeneous platform

Abstract

The invention discloses a gene data interpretation-oriented reconfigurable heterogeneous platform. The platform comprises a heterogeneous processor unit, an interconnection bus module, an internal memory, a gene interpretation data instruction input unit and a gene interpretation result output unit, wherein the heterogeneous processor unit is connected with each of the internal memory, the gene interpretation data instruction input unit and the gene interpretation result output unit through the interconnection bus module; the heterogeneous processor unit comprises a CPU, a GPU and an FPGA in which a DSP is integrated; the CPU forms a control engine; and the CPU, the GPU and the FPGA form an interpretation engine. According to the platform disclosed by the invention, the hardware support is provided for improving the accuracy and readability of gene data interpretation, and the platform has the advantages of being high in gene data interpretation efficiency, low in manufacturing cost and low in interpretation energy consumption.

Description

technical field [0001] The invention relates to gene sequencing technology, in particular to a reconfigurable heterogeneous platform oriented to gene data interpretation. Background technique [0002] In recent years, with the widespread application of Next Generation Sequencing technology (Next Generation Sequence, hereinafter referred to as NGS), the cost of gene sequencing has dropped rapidly, and gene technology has begun to be widely used. NGS includes two steps: genetic data calculation and genetic data interpretation. Genetic data calculation refers to the preprocessing of the original gene sequencing data, such as de-falsification and de-duplication, so that it can be used in the interpretation of genetic data. Genetic data interpretation refers to the analysis of genetic data. Analyze, reveal and explain the scientific meaning of genetic data processed by calculation in the fields of biology, medicine, health care and so on. [0003] At present, a bottleneck restri...

AI Technical Summary

Problems solved by technology

The disadvantage is: to achieve the same function, FPGA is generally slower than ASIC (Application Specific Integrated Circuit, ASIC), and the circuit area is larger than ASIC

However, there are two problems with the current GPU: First, due to the hardware structural characteristics of the GPU, many parallel algorithms cannot be effectively executed on the GPU; second, the GPU will generate a lot of heat during operation, and the energy consumption is high

However, in view of the aforementioned bottlenecks in the development of clinical application of gene technology, how to use the above-mentioned processors to build a hybrid architecture platform to realize the interpretation of massive genetic data has become a key technical problem to be solved urgently

Images