Title
A 7.3 M Output Non-Zeros/J Sparse Matrix-Matrix Multiplication Accelerator using Memory Reconfiguration in 40 nm
Abstract
A Sparse Matrix-Matrix multiplication (SpMM) accelerator with 48 heterogeneous cores and a reconfigurable memory hierarchy is fabricated in 40 nm CMOS. On-chip memories are reconfigured as scratchpad or cache and interconnected with synthesizable coalescing crossbars for efficient memory access in each phase of the algorithm. The <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$2.0\ \text{mm}\times 2.6\ \text{mm}$</tex> chip exhibits <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$12.6\times(8.4\times)$</tex> energy efficiency gain, <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$11.7\times(77.6\times)$</tex> off-chip bandwidth efficiency gain and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$17.1\times(36.9\times)$</tex> compute density gain against a high-end CPU (GPU) across a diverse set of synthetic and real-world power-law graph based sparse matrices.
Year
DOI
Venue
2019
10.23919/VLSIT.2019.8776507
2019 Symposium on VLSI Technology
Keywords
Field
DocType
Sparse matrix multiplier,synthesizable crossbar,decoupled access-execution,reconfigurablility and accelerator
Topology,Central processing unit,System on a chip,Memory hierarchy,Cache,Computer science,Parallel computing,Chip,Multiplication,Matrix multiplication,Sparse matrix
Conference
ISSN
ISBN
Citations 
0743-1562
978-1-7281-1530-6
0
PageRank 
References 
Authors
0.34
0
18
Name
Order
Citations
PageRank
subhankar pal1325.27
Dong-Hyeon Park2282.49
Siying Feng3102.25
Paul Gao4162.41
Jielun Tan532.41
Austin Rovinski61377.24
Shaolin Xie700.68
Chun Zhao8277.67
Aporva Amarnath9395.18
Timothy Wesley1000.34
Jonathan Beaumont11362.85
Kuan-Yu Chen1245055.78
Chaitali Chakrabarti131978184.17
Michael Bedford Taylor141707154.51
Trevor Mudge156139659.74
David Blaauw168916823.47
Hun-Seok Kim1729427.15
Ronald G. Dreslinski18125881.02