AxE: An Approximate-Exact Multi-Processor System-on-Chip Platform
A.S.
Baroughi
School of Electrical Engineering
Iran University of Science and Technology
Tehran, Iran |
S. Huemer
Technische Universitat Wien, Vienna, Austria |
H. S. Shahhoseini
School of Electrical Engineering
Iran University of Science and Technology
Tehran, Iran |
N. TaheriNejad
TU Wien
Vienna, Austria |
PDF │ Abstract │ Keywords │ References │ Cite This
Abstract:
Due to the ever-increasing complexity of computing tasks, emerging computing paradigms that increase efficiency, such as approximate computing, are gaining momentum. However, so far, the majority of proposed solutions for hardware-based approximation have been application-specific and/or limited to smaller units of the computing system and require engineering effort for integration into the rest of the system. In this paper, we present Approximate and Exact Multi-Processor system-on-chip (AxE) platform. AxE is the first general-purpose approximate Multi-Processor System-on-Chip (MPSoC). AxE is a heterogeneous RISC-V platform with exact and approximate cores that allows exploring hardware approximation for any application and using software instructions. Using the full capacity of an entire MPSoC, especially a heterogeneous one such as AxE, is an increasingly challenging problem. Therefore, we also propose a task mapping method for running exact and approximable applications on AxE. That is a mixed task mapping, in which applications are viewed as a set of tasks that can be run independently on different processors with different capabilities (exact or approximate). We evaluated our proposed method on AxE and reached a 32% average execution speed-up and 21% energy consumption saving with an average of 99.3% accuracy on three mixed workloads. We also ran a sample image processing application, namely gray-scale filter, on AxE and will present its results.
Keywords: Approximation Computing; Multi-Processor System-on-Chip (MPSoC); Approximate and Exact MPSoC; Task Mapping; RISC-V
References:
[1] I. Elsadek and E. Y. Tawfik, "RISC-V resource-constrained cores: A survey and energy comparison", 2021 19th IEEE International New Circuits and Systems Conference (NEWCAS), pp. 1-5, 2021.
[2] E. Shamsa, A. Kanduri, N. TaheriNejad, A. Proebstl Chakraborty, A. M. Rahmani and P. Liljeberg, "User-centric resource management for embedded multi-core processors", The 33rd International Conference on VLSI Design and The 19th International Conference on Embedded Design, pp. 1-6, 2020.
[3] E. Shamsa, A. Proebstl, N. TaheriNejad, A. Kanduri, A. M. Rahmani, P. Liljeberg, et al., "Ubar: User and battery aware resource management for smartphones", ACM Transactions on Embedded Computing Systems (TECS), pp. 1-23, 2021.
[4] S. Sidiroglou-Douskos, S. Misailovic, H. Hoffmann and M. Rinard, "Managing performance vs. accuracy trade-offs with loop perforation", Proceedings of the 19th ACM SIGSOFT symposium and the 13th European conference on Foundations of software engineering, pp. 124-134, 2011.
[5] S. Misailovic, S. Sidiroglou, H. Hoffmann and M. Rinard, "Quality of service profiling", 2010 ACM/IEEE 32nd International Conference on Software Engineering, pp. 25-34, 2010.
[6] A. K. Mishra, R. Barik and S. Paul, iACT: A software-hardware framework for understanding the scope of approximate computing, 2014.
[7] Y. Kim, S. Venkataramani, S. Sen and A. Raghunathan, "Value similarity extensions for approximate computing in general-purpose processors", 2021 Design Automation Test in Europe Conference Exhibition (DATE), pp. 481-486, 2021.
[8] G. Tziantzioulis, N. Hardavellas and S. Campanoni, "Temporal approximate function memorization", IEEE Micro, vol. 38, no. 4, pp. 60-70, 2018.
[9] J. Lee, H. Seo, H. Seok and Y. Kim, "A novel approximate adder design using error reduced carry prediction and constant truncation", IEEE Access, vol. 9, pp. 119939-119953, 2021.
[10] S. E. Fatemieh, M. R. Reshadinezhad and N. TaheriNejad, "Approximate in-memory computing using memristive imply logic and its application to image processing", 2022 IEEE International Symposium on Circuits and Systems (IS CAS), pp. 1-5, 2022.
[11] P. J. Edavoor, S. Raveendran and A. D. Rahulkar, "Approximate multiplier design using novel dual-stage 4:2 compressors", IEEE Access, vol. 8, pp. 48337-48351, 2020.
[12] G. Anusha and P. Deepa, "Design of approximate adders and multipliers for error tolerant image processing", Microprocessors and Microsystems, vol. 72, pp. 102940, 2020.
[13] H. Seok, H. Seo, J. Lee and Y. Kim, "COREA: Delay- and energy-efficient approximate adder using effective carry speculation", Electronics, vol. 10, no. 18, 2021.
[14] S. Ullah, S. Rehman, B. S. Prabakaran, F. Kriebel, M. A. Hanif, M. Shafique, et al., "Area-optimized low-latency approximate multipliers for FPGA-based hardware accelerators", Proceedings of the 55th Annual Design Automation Conference ser. DAC '18, 2018.
[15] S. Vahdat, M. Kamal, A. Afzali-Kusha and M. Pedram, "TOSAM: An energy-efficient truncation- and rounding-based scalable approximate multiplier", IEEE Transactions on Very Large Scale Integration (VLSI) Systems, vol. 27, no. 5, pp. 1161-1173, 2019.
[16] R. Nayar, P. Balasubramanian and D. L. Maskell, "Hardware optimized approximate adder with normal error distribution", 2020 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), pp. 84-89, July 2020.
[17] Y. Chen, X. Yang, F. Qiao, J. Han, Q. Wei and H. Yang, "A multi-accuracy-level approximate memory architecture based on data significance analysis", 2016 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), pp. 385-390, 2016.
[18] M. Kang, S. K. Gonugondla and N. R. Shanbhag, "Deep in-memory architectures in SRAM: An analog approach to approximate computing", Proceedings of the IEEE, vol. 108, no. 12, pp. 2251-2275, 2020.
[19] T. Trevisan et al., "Approxrisc: An approximate computing infrastructure for RISC-V", RISC-V Workshop in Barcelona, May 2018.
[20] N. A. Said et al., "FPU bit-width optimization for approximate computing: A non-intrusive approach", 2020 15th Design Technology of Integrated Systems in Nanoscale Era (DTIS), pp. 1-6, 2020.
[21] I. Felzmann, J. F. Filho and L. Wanner, "Risk-5: Controlled approximations for RISC-V", IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, vol. 39, no. 11, pp. 4052-4063, 2020.
[22] C. Wolf, PicoRV32 - A Size-Optimized RISC-V CPU, [online] Available: https://github.com/cliffordwolf/picorv32.
[23] V. Mrazek et al., "Evoapprox8b: Library of approximate adders and multipliers for circuit design and benchmarking of approximation methods", DATE 2017, pp. 258-261, 2017.
[24] M. K. Papamichael and J. C. Hoe, Connect: Re-examining conventional wisdom for designing NOCs in the context of FPGAs, Association for Computing Machinery, 2012.
[25] A. De Myttenaere, B. Golden, B. Le Grand and F. Rossi, "Mean absolute percentage error for regression models", Neurocomputing, vol. 192, pp. 38-48, 2016.
Cite this paper as:
A. S. Baroughi, S. Huemer, H. S. Shahhoseini and N. TaheriNejad, "AxE: An Approximate-Exact Multi-Processor System-on-Chip Platform," 2022 25th Euromicro Conference on Digital System Design (DSD), Maspalomas, Spain, 2022, pp. 60-66, doi: 10.1109/DSD57027.2022.00018.
|