Abstract
In this chapter, we discuss hardware–software co-design approaches that enable machine learning inference on ultra-resource-constrained embedded systems. As our use case, we consider the printed electronics. The latter form an extreme example of embedded machine learning application. Printed electronics constitute a promising solution to bring computing and smart services in application domains that require sub-cent cost and conformality and have not seen yet significant penetration of computing. Printed electronics form a rapidly growing market but also feature several prevalent limitations. Integration density and performance of printed electronics are of order of magnitude lower than those in silicon VLSI systems, and implementing complex circuits, e.g., ML classifiers, poses a great challenge. Considering the a priori requirement of embedded ML for acceptable accuracy and low latency within the limitations of the ultra-resource-constrained printed devices, custom-designed circuits as well as software–hardware co-design and optimization combined with non-conventional computing approaches (e.g., approximate and stochastic computing) are becoming mandatory for the realization of such circuits.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Alaghi, A., Hayes, J.P.: Survey of stochastic computing. ACM Trans. Embed. Comput. Syst. 12(2s), (2013). http://dx.doi.org/10.1145/2465787.2465794
Armeniakos, G., Zervakis, G., Soudris, D., Tahoori, M.B., Henkel, J.: Cross-layer approximation for printed machine learning circuits. In: Design, Automation Test in Europe Conference & Exhibition (DATE) (2022)
Ayub, M.K., Hasan, O., Shafique, M.: Statistical error analysis for low power approximate adders. In: Design Automation Conference (DAC), pp 1–6 (2017)
Balaskas, K., Zervakis, G., Amrouch, H., Henkel, J., Siozios, K.: Automated design approximation to overcome circuit aging. IEEE Trans. Circ. Syst. I Reg. Pap. 68(11), 4710–4721 (2021)
Balaskas, K., Zervakis, G., Siozios, K., Tahoori, M.B., Henkel, J.: Approximate decision trees for machine learning classification on tiny printed circuits. In: International Symposium on Quality Electronic Design (ISQED) (2022)
Bhardwaj, K., Mane, P.S., Henkel, J.: Power- and area-efficient approximate Wallace tree multiplier for error-resilient systems. In: Fifteenth International Symposium on Quality Electronic Design, pp 263–269 (2014)
Biggs, J., Myers, J., Kufel, J., Özer, E., Craske, S., Sou, A., Ramsdale, C., Williamson, K., Price, R., White, S.: A natively flexible 32-bit arm microprocessor. Nature 595, 532–536 (2021)
Bleier, N., Mubarik, M., Rasheed, F., Aghassi-Hagmann, J., Tahoori, M.B., Kumar, R.: Printed microprocessors. In: Annu. Int. Symp. Computer Architecture (ISCA), pp. 213–226 (2020)
Chang, J.S., Facchetti, A.F., Reuss, R.: A circuits and systems perspective of organic/printed electronics: Review, challenges, and contemporary and emerging design approaches. IEEE J. Emerg. Sel. Top. Circ. Syst. 7(1), 7–26 (2017). https://doi.org/10.1109/JETCAS.2017.2673863
Cherupalli, H., Duwe, H., Ye, W., Kumar, R., Sartori, J.: Bespoke processors for applications with ultra-low area and power constraints. In: Annu. Int. Symp. Computer Architecture (ISCA), pp 41–54 (2017)
Choi, J., Venkataramani, S.: Approximate computing techniques for deep neural networks. In: Reda, S., Shafique, M. (eds.) Approximate Circuits: Methodologies and CAD, Springer International Publishing, Cham, pp 307–329 (2019)
Chung, S., Kim, S.O., Kwon, S.K., Lee, C., Hong, Y.: All-inkjet-printed organic thin-film transistor inverter on flexible plastic substrate. IEEE Electron Dev. Lett. 32(8), 1134–1136 (2011)
Conti, S., Pimpolari, L., Calabrese, G., Worsley, R., Majee, S., Polyushkin, D.K., Paur, M., Pace, S., Keum, D.H., Fabbri, F., et al.: Low-voltage 2D materials-based printed field-effect transistors for integrated digital and analog electronics on paper. Nature Communications 11(1):1–9 (2020)
Cui, Z.: Printed Electronics: Materials, Technologies and Applications. Wiley (2016)
Douthwaite, M., Garcıa-Redondo, F., Georgiou, P., Das, S.: A time-domain current-mode MAC engine for analogue neural networks in flexible electronics. In: IEEE Biomedical Circuits and Systems Conference (BioCAS), pp. 1–4 (2019)
Dua, D., Graff, C.: UCI machine learning repository (2017)
Erozan, A.T., Wang, G.Y., Bishnoi, R., Aghassi-Hagmann, J., Tahoori, M.B.: A compact low-voltage true random number generator based on inkjet printing technology. IEEE Trans. Very Large Scale Integr. (VLSI) Syst. 28(6), 1485–1495 (2020)
Hanif, M.A., Khalid, F., Shafique, M.: CANN: Curable approximations for high-performance deep neural network accelerators. In: Design Automation Conference (DAC) (2019)
Hayes, J.P.: Introduction to stochastic computing and its challenges. In: Design Automation Conference (DAC) (2015)
Huber, B., Popp, P., Kaiser, M., Ruediger, A., Schindler, C.: Fully inkjet printed flexible resistive memory. Appl. Phys. Lett. 110(14), 143503 (2017)
Kim, J., Jeerapan, I., Imani, S., Cho, T.N., Bandodkar, A., Cinti, S., Mercier, P.P., Wang, J.: Noninvasive alcohol monitoring using a wearable tattoo-based iontophoretic-biosensing system. ACS Sensors 1(8), 1011–1019 (2016)
Kondo, M., Uemura, T., Akiyama, M., Namba, N., Sugiyama, M., Noda, Y., Araki, T., Yoshimoto, S., Sekitani, T.: Design of ultraflexible organic differential amplifier circuits for wearable sensor technologies. In: 2018 IEEE International Conference on Microelectronic Test Structures (ICMTS), pp. 79–84. IEEE (2018)
Liu, Y., Liu, S., Wang, Y., Lombardi, F., Han, J.: A survey of stochastic computing neural networks for machine learning applications. IEEE Trans. Neural Networks Learn. Syst. 32(7), 2809–2824 (2021)
Mostafalu, P., Lenk, W., Dokmeci, M.R., Ziaie, B., Khademhosseini, A., Sonkusale, S.R.: Wireless flexible smart bandage for continuous monitoring of wound oxygenation. IEEE Trans. Biomed. Circ. Syst. 9(5), 670–677 (2015)
Mrazek, V., Hanif, M.A., Vasicek, Z., Sekanina, L., Shafique, M.: autoAX: An automatic design space exploration and circuit building methodology utilizing libraries of approximate components. In: Design Automation Conference (2019a)
Mrazek, V., Vasicek, Z., Sekanina, L., Hanif, M.A., Shafique, M.: ALWANN: Automatic layer-wise approximation of deep neural network accelerators without retraining. In: Int Conference on Computer-Aided Design (ICCAD) (2019b)
Mubarik, M.H., Weller, D.D., Bleier, N., Tomei, M., Aghassi-Hagmann, J., Tahoori, M.B., Kumar, R.: Printed machine learning classifiers. In: Annu. Int. Symp. Microarchitecture (MICRO), pp. 73–87 (2020)
Ozer, E., Kufel, J., Biggs, J., Brown, G., Myers, J., Rana, A., Sou, A., Ramsdale, C.: Bespoke machine learning processor development framework on flexible substrates. In: 2019 IEEE International Conference on Flexible and Printable Sensors and Systems (FLEPS), pp. 1–3 (2019)
Prakash, S., Callahan, T., Bushagour, J., Banbury, C.R., Green, A.V., Warden, P., Ansell, T., Reddi, V.J.: CFU playground: Full-stack open-source framework for tiny machine learning (TinyML) acceleration on FPGAs (2022). CoRR abs/2201.01863. https://arxiv.org/abs/2201.01863
Ren, A., Li, Z., Ding, C., Qiu, Q., Wang, Y., Li, J., Qian, X., Yuan, B.: SC-DCNN: Highly-scalable deep convolutional neural network using stochastic computing. In: Proceedings of the Twenty-Second International Conference on Architectural Support for Programming Languages and Operating Systems, pp. 405–418 (2017)
Shafique, M., Ahmad, W., Hafiz, R., Henkel, J.: A low latency generic accuracy configurable adder. In: Design Automation Conference (DAC), pp. 1–6 (2015)
Shafique, M., Theocharides, T., Reddy, V.J., Murmann, B.: TinyML: Current progress, research challenges, and future roadmap. In: Design Automation Conference (DAC), pp. 1303–1306 (2021). https://doi.org/10.1109/DAC18074.2021.9586232
Shao, F., Wan, Q.: Recent progress on jet printing of oxide-based thin film transistors. J. Phys. D Appl. Phys. 52(14), 143002 (2019)
Spantidi, O., Zervakis, G., Anagnostopoulos, I., Amrouch, H., Henkel, J.: Positive/negative approximate multipliers for DNN accelerators. In: International Conference on Computer Aided Design (ICCAD), pp. 1–9 (2021)
Tasoulas, Z.G., Zervakis, G., Anagnostopoulos, I., Amrouch, H., Henkel, J.: Weight-oriented approximation for energy-efficient neural network inference accelerators. IEEE Trans. Circ. Syst. I Reg. Papers 67, 4670–4683 (2020)
Weller, D., Marques, G.C., Aghassi-Hagmann, J., Tahoori, M.B.: An inkjet-printed low-voltage latch based on inorganic electrolyte-gated transistors. IEEE Electron Dev. Lett. 39(6), 831–834 (2018)
Weller, D.D., Hefenbrock, M., Tahoori, M.B., Aghassi-Hagmann, J., Beigl, M.: Programmable neuromorphic circuit based on printed electrolyte-gated transistors. In: 2020 25th Asia and South Pacific Design Automation Conference (ASP-DAC), pp. 446–451 (2020). https://doi.org/10.1109/ASP-DAC47756.2020.9045211
Weller, D.D., Bleier, N., Hefenbrock, M., Aghassi-Hagmann, J., Beigl, M., Kumar, R., Tahoori, M.B.: Printed stochastic computing neural networks. In: Design, Automation Test in Europe Conference Exhibition (DATE), pp. 914–919 (2021)
Özer, E., Kufel, J., Myers, J., Biggs, J., Brown, G., Rana, A., Sou, A., Ramsdale, C., White, S.: A hardwired machine learning processing engine fabricated with submicron metal-oxide thin-film transistors on a flexible substrate. Nature Electronics 3, 1–7 (2020)
Zervakis, G., Tsoumanis, K., Xydis, S., Soudris, D., Pekmestzi, K.: Design-efficient approximate multiplication circuits through partial product perforation. IEEE Trans. Very Large Scale Integr. (VLSI) Syst. 24(10), 3105–3117 (2016)
Zervakis, G., Ntouskas, F., Xydis, S., Soudris, D., Pekmestzi, K.: VOSsim: A framework for enabling fast voltage overscaling simulation for approximate computing circuits. IEEE Trans. Very Large Scale Integr. (VLSI) Syst. 26(6), 1204–1208 (2018)
Zervakis, G., Koliogeorgi, K., Anagnostos, D., Zompakis, N., Siozios, K.: VADER: Voltage-driven netlist pruning for cross-layer approximate arithmetic circuits. IEEE Trans. Very Large Scale Integr. (VLSI) Syst. 27(6), 1460–1464 (2019)
Zervakis, G., Xydis, S., Soudris, D., Pekmestzi, K.: Multi-level approximate accelerator synthesis under voltage island constraints. IEEE Trans. Circ. Syst. II Exp. Briefs 66(4), 607–611 (2019)
Zervakis, G., Amrouch, H., Henkel, J.: Design automation of approximate circuits with runtime reconfigurable accuracy. IEEE Access 8, 53522–53538 (2020)
Zervakis, G., Saadat, H., Amrouch, H., Gerstlauer, A., Parameswaran, S., Henkel, J.: Approximate computing for ML: State-of-the-art, challenges and visions. In: Asia and South Pacific Design Automation Conference, pp. 189–196 (2021a)
Zervakis, G., Spantidi, O., Anagnostopoulos, I., Amrouch, H., Henkel, J.: Control variate approximation for DNN accelerators. In: Design Automation Conference (DAC), pp. 481–486 (2021b)
Acknowledgements
This work is partially supported by the German Research Foundation (DFG) through the project “ACCROSS: Approximate Computing aCROss the System Stack” HE 2343/16-1 and by grant from the Excellence Initiative of Karlsruhe Institute of Technology under Future Field program “SoftNeuro.”
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2024 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this chapter
Cite this chapter
Zervakis, G., Tahoori, M.B., Henkel, J. (2024). Hardware–Software Co-design for Ultra-Resource-Constrained Embedded Machine Learning Inference: A Printed Electronics Use Case. In: Pasricha, S., Shafique, M. (eds) Embedded Machine Learning for Cyber-Physical, IoT, and Edge Computing. Springer, Cham. https://doi.org/10.1007/978-3-031-39932-9_8
Download citation
DOI: https://doi.org/10.1007/978-3-031-39932-9_8
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-39931-2
Online ISBN: 978-3-031-39932-9
eBook Packages: EngineeringEngineering (R0)