Multipath data floating point processor prototype

Abstract

The invention discloses a multipath data floating point processor prototype. A framework is characterized by aligning structured data, half-structured data and non-structured data to be three paths of arrays; applying a high singular value decomposition, and decomposing the three arrays to be a two-order tensor matrix mode; transforming the matrix mode to a sparse domain, and performing blocked floating point quantization treatment. Finally, the multipath data floating point processor prototype is structured.

Description

Technical field [0001] The invention relates to a multi-channel data floating point processor prototype Background technique [0002] With cloud computing and virtualization, showing the characteristics of "large scale", "high density", "high energy consumption", and "complexity", the construction and development of a new generation of data centers and the improvement of data center infrastructure management will become increasingly Importantly, the integrated management and intelligence of the data center infrastructure will become a new trend in the development of data centers. The super-large data center provides the entire application service from the infrastructure to the subsequent data analysis, screening, and application. Not only data analysis, but also cloud computing that specializes in intelligent manufacturing, which is different from the generalized services provided by public clouds, and super computing, which puts forward higher requirements for big data processi...

AI Technical Summary

Problems solved by technology

[0008] Using block floating-point arithmetic divides the data into groups that are scaled relative to each other, but not to the same scale as members of other groups, even for simple mathematical operations such as multiplication

Images