High-efficiency Compressor Trees for Latest AMD FPGAs
Open Access
- 30 April 2024
- journal article
- research article
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Reconfigurable Technology and Systems
- Vol. 17 (2) , 1-32
Abstract
High-fan-in dot product computations are ubiquitous in highly relevant application domains, such as signal processing and machine learning. Particularly, the diverse set of data formats used in machine learning poses a challenge for flexible efficient design solutions. Ideally, a dot product summation is composed from a carry-free compressor tree followed by a terminal carry-propagate addition. On FPGA, these compressor trees are constructed from generalized parallel counters whose architecture is closely tied to the underlying reconfigurable fabric. This work reviews known counter designs and proposes new ones in the context of the new AMD Versal™ fabric. On this basis, we develop a compressor generator featuring variable-sized counters, novel counter composition heuristics, explicit clustering strategies, and case-specific optimizations like logic gate absorption. In comparison to the Vivado™ default implementation, the combination of such a compressor with a novel, highly efficient quaternary adder reduces the LUT footprint across different bit matrix input shapes by 45% for a plain summation and by 46% for a terminal accumulation at a slight cost in critical path delay still allowing an operation well above 500 MHz. We demonstrate the aptness of our solution at examples of low-precision integer dot product accumulation units.Keywords
Funding Information
- European Union's Horizon 2020 Research and Innovation Program under the Marie Skłodowska Curie (956090)
This publication has 20 references indexed in Scilit:
- XOR-Net: An Efficient Computation Pipeline for Binary Neural Network Inference on Edge Devices2020
- Area Optimized Synthesis of Compressor Trees on Xilinx FPGAs Using Generalized Parallel CountersIEEE Access, 2019
- Fast inference of deep neural networks in FPGAs for particle physicsJournal of Instrumentation, 2018
- Advanced Compressor Tree Synthesis for FPGAsIEEE Transactions on Computers, 2018
- XNOR-Net: ImageNet Classification Using Binary Convolutional Neural Networks2016
- Pipelined compressor tree optimization using integer linear programming2014
- Arithmetic core generation using bit heaps2013
- Exploiting fast carry-chains of FPGAs for designing compressor trees2009
- Efficient synthesis of compressor trees on FPGAs2008
- A Suggestion for a Fast MultiplierIEEE Transactions on Electronic Computers, 1964