Throughput-Distortion Computation of Generic Matrix Multiplication: Toward a Computation Channel for Digital Signal Processing Systems


The generic matrix multiply (GEMM) function is the core element of high-performance linear algebra libraries used in many computationally demanding digital signal processing (DSP) systems. We propose an acceleration technique for GEMM based on dynamically adjusting the imprecision (distortion) of computation. Our technique employs adaptive scalar companding and rounding to input matrix blocks followed by two forms of packing in floating-point that allow for concurrent calculation of multiple results. Since the adaptive companding process controls the increase of concurrency (via packing), the increase in processing throughput (and the corresponding increase in distortion) depends on the input data statistics. To demonstrate this, we derive the optimal throughput-distortion control framework for GEMM for the broad class of zero-mean, independent identically distributed, input sources. Our approach converts matrix multiplication in programmable processors into a computation channel: when increasing the processing throughput, the output noise (error) increases due to: (i) coarser quantization; and (ii) computational errors caused by exceeding the machine-precision limitations. We show that, under certain distortion in the GEMM computation, the proposed framework can significantly surpass 100% of the peak performance of a given processor. The practical benefits of our proposal are shown in a face recognition system and a multilayer perceptron system trained for metadata learning from a large music feature database.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here

PROJECT TITLE : Secure and Efficient Data Transmission for Cluster-Based Wireless Sensor Networks - 2014 ABSTRACT: Secure data transmission is a critical issue for wireless sensor networks (WSNs). Clustering is an effective
PROJECT TITLE : Multi-Core Embedded Wireless Sensor Networks Architecture and Applications - 2014 ABSTRACT: Technological advancements in the silicon industry, as predicted by Moore's law, have enabled integration of billions
PROJECT TITLE : Multicast Capacity in MANET with Infrastructure Support - 2014 ABSTRACT: We study the multicast capacity under a network model featuring both node's mobility and infrastructure support. Combinations between
PROJECT TITLE : Joint Routing and Resource Allocation for Delay Minimization in Cognitive Radio Based Mesh Networks - 2014 ABSTRACT: This paper studies the joint design of routing and resource allocation algorithms in cognitive
PROJECT TITLE : Efficient Data Collection for Large-Scale Mobile Monitoring Applications - 2014 ABSTRACT: Radio frequency identification (RFID) and wireless sensor networks (WSNs) have been popular in the industrial field,

Ready to Complete Your Academic MTech Project Work In Affordable Price ?

Project Enquiry