ABSTRACT
The Software Defined Architectures (SODA) Synthesizer is an open-source compiler-based tool able to automatically generate domain-specialized systems targeting Application-Specific Integrated Circuits (ASICs) or Field Programmable Gate Arrays (FPGAs) starting from high-level programming. SODA is composed of a frontend, SODA-OPT, which leverages the multilevel intermediate representation (MLIR) framework to interface with productive programming tools (e.g., machine learning frameworks), identify kernels suitable for acceleration, and perform high-level optimizations, and of a state-of-the-art high-level synthesis backend, Bambu from the PandA framework, to generate custom accelerators. One specific application of the SODA Synthesizer is the generation of accelerators to enable ultra-low latency inference and control on autonomous systems for scientific discovery (e.g., electron microscopes, sensors in particle accelerators, etc.). This paper provides an overview of the flow in the context of the generation of accelerators for edge processing to be integrated in transmission electron microscopy (TEM) devices, focusing on use cases from precision material synthesis. We show the tool in action with an example of design space exploration for inference on reconfigurable devices with a conventional deep neural network model (LeNet). Finally, we discuss the research directions and opportunities enabled by SODA in the area of autonomous control for scientific experimental workflows.
- E. Bethel and eds. 2016. Report of the DOE Workshop on Management, Analysis, and Visualization of Experimental and Observational data - The Convergence of Data and Computing. Technical Report.Google Scholar
- Nicolas Bohm Agostini, Serena Curzel, Vinay Amatya, Cheng Tan, Marco Minutoli, Vito Giovanni Castellana, Joseph Manzano, David Kaeli, and Antonino Tumeo. 2022. An MLIR-based Compiler Flow for System-Level Design and Hardware Acceleration. In 41st IEEE/ACM International Conference on Computer-Aided Design (ICCAD'22). To appear.Google Scholar
- Nicolas Bohm Agostini, Serena Curzel, Jeff Zhang, Ankur Limaye, Cheng Tan, Vinay Amatya, Marco Minutoli, Vito Giovanni Castellana, Joseph Manzano, David Brooks, Gu-Yeon Wei, and Antonino Tumeo. 2022. Bridging Python to Silicon: The SODA Toolchain. IEEE Micro (2022). Google ScholarCross Ref
- Vito Giovanni Castellana and Fabrizio Ferrandi. 2013. An automated flow for the High Level Synthesis of coarse grained parallel applications. In 2013 International Conference on Field-Programmable Technology (FPT). 294--301. Google ScholarCross Ref
- Vito Giovanni Castellana, Antonino Tumeo, and Fabrizio Ferrandi. 2021. High-Level Synthesis of Parallel Specifications Coupling Static and Dynamic Controllers. In IEEE International Parallel and Distributed Processing Symposium (IPDPS'21). 192--202. Google ScholarCross Ref
- S. Curzel, N. Bohm Agostini, S. Song, I. Dagli, A. Limaye, M. Minutoli, V. G. Castellana, V. Amatya, J. Manzano, A. Das, F. Ferrandi, and A. Tumeo. 2021. Automated Generation of Integrated Digital and Spiking Neuromorphic Machine Learning Accelerators. In ICCAD: International Conference On Computer Aided Design. 1--7.Google Scholar
- J. Duarte, S. Han, P. Harris, S. Jindariani, E. Kreinar, B. Kreis, J. Ngadiuba, M. Pierini, R. Rivera, N. Tran, and Z. Wu. 2018. Fast inference of deep neural networks in FPGAs for particle physics. Journal of Instrumentation 13, 07 (jul 2018), P07027--P07027. Google ScholarCross Ref
- Fabrizio Ferrandi, Vito Giovanni Castellana, Serena Curzel, Pietro Fezzardi, Michele Fiorito, Marco Lattuada, Marco Minutoli, Christian Pilato, and Antonino Tumeo. 2021. Bambu: an Open-Source Research Framework for the High-Level Synthesis of Complex Applications. In 58th ACM/IEEE Design Automation Conference (DAC'21). 1327--1330. Google ScholarDigital Library
- Andrew B. Kahng and Tom Spyrou. 2021. The OpenROAD Project: Unleashing Hardware Innovation. In Government Microcircuit Applications and Critical Technology Conference. 1--6.Google Scholar
- Emir Kocer, Tsz Wai Ko, and Jörg Behler. 2021. Neural Network Potentials: A Concise Overview of Methods. Google ScholarCross Ref
- Chris Lattner, Mehdi Amini, Uday Bondhugula, Albert Cohen, Andy Davis, Jacques Pienaar, River Riddle, Tatiana Shpeisman, Nicolas Vasilache, and Oleksandr Zinenko. 2021. MLIR: Scaling Compiler Infrastructure for Domain Specific Computation. In IEEE/ACM International Symposium on Code Generation and Optimization (CGO'21). 2--14. Google ScholarDigital Library
- Y. Lecun, L. Bottou, Y. Bengio, and P. Haffner. 1998. Gradient-based learning applied to document recognition. Proc. IEEE 86, 11 (1998), 2278--2324. Google ScholarCross Ref
- Haotong Liang, Valentin Stanev, Aaron Gilad Kusne, Yuto Tsukahara, Kaito Ito, Ryota Takahashi, Mikk Lippmaa, and Ichiro Takeuchi. 2022. Application of machine learning to reflection high-energy electron diffraction images for automated structural phase mapping. Phys. Rev. Materials 6 (Jun 2022), 063805. Issue 6.Google ScholarCross Ref
- John E. Mahan, Kent M. Geib, G. Y. Robinson, and Robert G. Long. 1990. A review of the geometrical fundamentals of reflection high-energy electron diffraction with application to silicon surfaces. Journal of Vacuum Science & Technology A 8, 5 (1990), 3692--3700.Google ScholarCross Ref
- Paolo Mantovani, Davide Giri, Giuseppe Di Guglielmo, Luca Piccolboni, Joseph Zuckerman, Emilio G. Cota, Michele Petracca, Christian Pilato, and Luca P. Carloni. 2020. Agile SoC Development with Open ESP. In IEEE/ACM International Conference On Computer Aided Design (ICCAD'20). 1--9. Google ScholarDigital Library
- Marco Minutoli, Vito Giovanni Castellana, Nicola Saporetti, Stefano Devecchi, Marco Lattuada, Pietro Fezzardi, Antonino Tumeo, and Fabrizio Ferrandi. 2022. Svelto: High-Level Synthesis of Multi-Threaded Accelerators for Graph Analytics. IEEE Trans. Comput. 71, 3 (March 2022), 520--533. Google ScholarDigital Library
- Marco Minutoli, Vito Giovanni Castellana, Antonino Tumeo, and Fabrizio Ferrandi. 2015. Inter-procedural resource sharing in High Level Synthesis through function proxies. In 25th International Conference on Field Programmable Logic and Applications (FPL'15). 1--8. Google ScholarCross Ref
- Cassandra Pate, James Hart, and Mitra Taheri. 2021. RapidEELS: machine learning for denoising and classification in rapid acquisition electron energy loss spectroscopy. Scientific Reports 11 (09 2021). Google ScholarCross Ref
- Sydney R. Provence, Suresh Thapa, Rajendra Paudel, Tristan K. Truttmann, Abhinav Prakash, Bharat Jalan, and Ryan B. Comes. 2020. Machine learning analysis of perovskite oxides grown by molecular beam epitaxy. Physical Review Materials 4, 8 (aug 2020). Google ScholarCross Ref
- W. Snyder. 2022. RapidEELS. https://github.com/patecm/rapidEELS Online, accessed 11-2022.Google Scholar
- Xifan Tang, Edouard Giacomin, Aurélien Alacchi, Baudouin Chauviere, and Pierre-Emmanuel Gaillardon. 2019. OpenFPGA: An Opensource Framework Enabling Rapid Prototyping of Customizable FPGAs. In 29th International Conference on Field Programmable Logic and Applications (FPL'19). 367--374. Google ScholarCross Ref
- Rama K. Vasudevan, Alexander Tselev, Arthur P. Baddorf, and Sergei V. Kalinin. 2014. Big-Data Reflection High Energy Electron Diffraction Analysis for Understanding Epitaxial Film Growth Processes. ACS Nano 8, 10 (2014), 10899--10908.Google ScholarCross Ref
- Ruizhe Zhao, Jianyi Cheng, Wayne Luk, and George A. Constantinides. 2022. POLSCA: Polyhedral High-Level Synthesis with Compiler Transformations. In 32nd International Conference on Field Programmable Logic and Applications (FPL'22). To appear.Google Scholar
Index Terms
- Towards On-Chip Learning for Low Latency Reasoning with End-to-End Synthesis
Recommendations
Recent developments in high-level synthesis
We survey recent developments in high level synthesis technology for VLSI design. The need for higher-level design automation tools are discussed first. We then describe some basic techniques for various subtasks of high-level synthesis. Techniques that ...
Layout-driven RTL binding techniques for high-level synthesis
ISSS '96: Proceedings of the 9th international symposium on System synthesisThe importance of effective and efficient accounting of layout effects is well-established in high-level synthesis (HLS), since it allows more realistic exploration of the design space and the generation of solutions with predictable metrics. This ...
On the use of VHDL-based behavioral synthesis for telecom ASIC design
ISSS '95: Proceedings of the 8th international symposium on System synthesisAbstract: VHDL-based behavioral synthesis is appearing on the market but it still has to prove that it can have a significant impact. In the past, most applications for behavioral synthesis came from the DSP area and from the academic world. In contrast,...
Comments