Visible to the public Reduced-Complexity Singular Value Decomposition For Tucker Decomposition: Algorithm And Hardware

TitleReduced-Complexity Singular Value Decomposition For Tucker Decomposition: Algorithm And Hardware
Publication TypeConference Paper
Year of Publication2020
AuthorsHu, X., Deng, C., Yuan, B.
Conference NameICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
KeywordsBig Data, big data analysis, CMOS integrated circuits, CMOS technology, compositionality, cyber physical systems, Data analysis, Data-Intensive Applications, decomposition, hardware architecture, high-dimensional data processing, Iterative methods, iterative self-multiplication, machine learning, Metrics, power 37.6 mW, pubcrawl, reduced-complexity SVD scheme, resource-constrained hardware platforms, singular value decomposition, size 28.0 nm, SVD, tensor algorithms, tensors, Tucker decomposition
AbstractTensors, as the multidimensional generalization of matrices, are naturally suited for representing and processing high-dimensional data. To date, tensors have been widely adopted in various data-intensive applications, such as machine learning and big data analysis. However, due to the inherent large-size characteristics of tensors, tensor algorithms, as the approaches that synthesize, transform or decompose tensors, are very computation and storage expensive, thereby hindering the potential further adoptions of tensors in many application scenarios, especially on the resource-constrained hardware platforms. In this paper, we propose a reduced-complexity SVD (Singular Vector Decomposition) scheme, which serves as the key operation in Tucker decomposition. By using iterative self-multiplication, the proposed scheme can significantly reduce the storage and computational costs of SVD, thereby reducing the complexity of the overall process. Then, corresponding hardware architecture is developed with 28nm CMOS technology. Our synthesized design can achieve 102GOPS with 1.09 mm2 area and 37.6 mW power consumption, and thereby providing a promising solution for accelerating Tucker decomposition.
Citation Keyhu_reduced-complexity_2020