Please use this identifier to cite or link to this item:
Full metadata record
DC FieldValueLanguage
dc.contributor.authorZhu, Shienen_US
dc.contributor.authorDuong, Luan H. K.en_US
dc.contributor.authorLiu, Weichenen_US
dc.identifier.citationZhu, S., Duong, L. H. K. & Liu, W. (2022). TAB : unified and optimized ternary, binary and mixed-precision neural network inference on the edge. ACM Transactions On Embedded Computing Systems.
dc.description.abstractTernary Neural Networks (TNNs) and mixed-precision Ternary Binary Networks (TBNs) have demonstrated higher accuracy compared to Binary Neural Networks (BNNs) while providing fast, low-power and memory-efficient inference. Related works have improved the accuracy of TNNs and TBNs, but overlooked their optimizations on CPU and GPU platforms. First, there is no unified encoding for the binary and ternary values in TNNs and TBNs. Second, existing works store the 2-bit quantized data sequentially in 32/64-bit integers, resulting in bit-extraction overhead. Last, adopting standard 2-bit multiplications for ternary values leads to a complex computation pipeline, and efficient mixed-precision multiplication between ternary and binary values is unavailable. In this paper, we propose TAB as a unified and optimized inference method for ternary, binary and mixed-precision neural networks. TAB includes unified value representation, efficient data storage scheme, and novel bitwise dot product pipelines on CPU/GPU platforms. We adopt signed integers for consistent value representation across binary and ternary values. We introduce a bitwidth-last data format that stores the first and second bits of the ternary values separately to remove the bit extraction overhead. We design the ternary and binary bitwise dot product pipelines based on Gated-XOR using up to 40% fewer operations than State-Of-The-Art (SOTA) methods. Theoretical speedup analysis shows that our proposed TAB-TNN is 2.3X fast as the SOTA ternary method RTN, 9.8X fast as 8-bit integer quantization (INT8), and 39.4X fast as 32-bit full-precision convolution (FP32). Experiment results on CPU and GPU platforms show that our TAB-TNN has achieved up to 34.6X speedup and 16X storage size reduction compared with FP32 layers. TBN, Binary-activation Ternary-weight Network (BTN) and BNN in TAB are up to 40.7X, 56.2X and 72.2X fast as FP32. TAB-TNN is up to 70.1% faster and 12.8% more power-efficient than RTN on Darknet-19 while keeping the same accuracy. TAB is open source as a PyTorch Extension for easy integration with existing CNN models.en_US
dc.description.sponsorshipMinistry of Education (MOE)en_US
dc.description.sponsorshipNanyang Technological Universityen_US
dc.relation.ispartofACM Transactions on Embedded Computing Systemsen_US
dc.rights© 2022 Association for Computing Machinery. All rights reserved. This paper was published in ACM Transactions on Embedded Computing Systems and is made available with permission of Association for Computing Machinery.en_US
dc.subjectEngineering::Computer science and engineering::Computing methodologies::Artificial intelligenceen_US
dc.subjectEngineering::Computer science and engineering::Computer systems organization::Performance of systemsen_US
dc.titleTAB : unified and optimized ternary, binary and mixed-precision neural network inference on the edgeen_US
dc.typeJournal Articleen
dc.contributor.schoolSchool of Computer Science and Engineeringen_US
dc.description.versionSubmitted/Accepted versionen_US
dc.subject.keywordsEdge Computingen_US
dc.subject.keywordsTernary Neural Networksen_US
dc.subject.keywordsBinary Neural Networksen_US
dc.description.acknowledgementThis work is partially supported by the Ministry of Education, Singapore, under its Academic Research Fund Tier 2 (MOE2019-T2-1-071) and Tier 1 (MOE2019-T1-001-072), and partially supported by Nanyang Technological University, Singapore, under its NAP (M4082282) and SUG (M4082087).en_US
item.fulltextWith Fulltext-
Appears in Collections:SCSE Journal Articles
Files in This Item:
File Description SizeFormat 
TECS_2021_TAB_Accepted Version.pdfThe Accepted Version3.55 MBAdobe PDFView/Open

Page view(s)

Updated on Jun 30, 2022


Updated on Jun 30, 2022

Google ScholarTM




Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.