Abstract
Computing-in-memory with emerging non-volatile memory (nvCiM) is shown to be a promising candidate for accelerating deep neural networks (DNNs) with high energy efficiency. However, most non-volatile memory (NVM) devices suffer from reliability issues, resulting in a difference between actual data involved in the nvCiM computation and the weight value trained in the data center. Thus, models actually deployed on nvCiM platforms achieve lower accuracy than their counterparts trained on the conventional hardware (e.g., GPUs). In this chapter, we first offer a brief introduction to the opportunities and challenges of nvCiM DNN accelerators and then show the properties of different types of NVM devices. We then introduce the general architecture of nvCiM DNN accelerators. After that, we discuss the source of unreliability and how to efficiently model their impact. Finally, we introduce representative works that mitigate the impact of device variations.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Note that each element of the output are deeply co-related, not independent.
References
Alibart F, Gao L, Hoskins BD, Strukov DB (2012) High precision tuning of state for memristive devices by adaptable variation-tolerant algorithm. Nanotechnology 23(7):075201
Aziz A, Breyer ET, Chen A, Chen X, Datta S, Gupta SK, Hoffmann M, Hu XS, Ionescu A, Jerry M et al (2018) Computing with ferroelectric FETs: devices, models, systems, and applications. In: 2018 Design, automation & test in Europe conference & exhibition (DATE). IEEE, pp 1289–1298
Aziz A, Ghosh S, Datta S, Gupta SK (2016) Physics-based circuit-compatible spice model for ferroelectric transistors. IEEE Electron Device Lett 37(6):805–808
Beck A, Bednorz J, Gerber C, Rossel C, Widmer D (2000) Reproducible switching effect in thin oxide films for memory applications. Appl Phys Lett 77(1):139–141
Berger L (1996) Emission of spin waves by a magnetic multilayer traversed by a current. Phys Rev B 54(13):9353
Carboni R, Ambrogio S, Chen W, Siddik M, Harms J, Lyle A, Kula W, Sandhu G, Ielmini D (2016) Understanding cycling endurance in perpendicular spin-transfer torque (p-STT) magnetic memory. In: 2016 IEEE International electron devices meeting (IEDM). IEEE, pp 21–6
Chappert C, Fert A, Van Dau FN (2010) The emergence of spin electronics in data storage. Nanosci Technol Collect Rev Nat J 147–157
Chen WH, Dou C, Li KX, Lin WY, Li PY, Huang JH, Wang JH, Wei WC, Xue CX, Chiu YC et al (2019) CMOS-integrated memristive non-volatile computing-in-memory for AI edge processors. Nat Electron 2(9):420–428
Chen Y, Luo T, Liu S, Zhang S, He L, Wang J, Li L, Chen T, Xu Z, Sun N et al (2014) DaDianNao: a machine-learning supercomputer. In: 2014 47th Annual IEEE/ACM international symposium on microarchitecture. IEEE, pp 609–622
Chi P, Li S, Xu C, Zhang T, Zhao J, Liu Y, Wang Y, Xie Y (2016) PRIME: a novel processing-in-memory architecture for neural network computation in ReRAM-based main memory. ACM SIGARCH Comput Architect News 44(3):27–39
Choi BJ, Torrezan AC, Strachan JP, Kotula P, Lohn A, Marinella MJ, Li Z, Williams RS, Yang JJ (2016) High-speed and low-energy nitride memristors. Adv Funct Mater 26(29):5290–5296
Draper J, Chame J, Hall M, Steele C, Barrett T, LaCoss J, Granacki J, Shin J, Chen C, Kang CW et al (2002) The architecture of the diva processing-in-memory chip. In: Proceedings of the 16th international conference on supercomputing, pp 14–25
Farmahini-Farahani A, Ahn JH, Morrow K, Kim NS (2015) NDA: near-dram acceleration architecture leveraging commodity dram devices and standard memory modules. In: 2015 IEEE 21st International symposium on high performance computer architecture (HPCA). IEEE, pp 283–295
Feinberg B, Wang S, Ipek E (2018) Making memristive neural network accelerators reliable. In: 2018 IEEE International symposium on high performance computer architecture (HPCA). IEEE, pp 52–65
Gao D, Huang Q, Zhang L, Yin X, Li B, Schlichtmann U, Zhuo C (2021) Bayesian inference based robust computing on memristor crossbar. In: 2021 56th ACM/IEEE Design automation conference (DAC). IEEE, pp 1–6
Gao D, Reis D, Hu XS, Zhuo C (2019) Eva-CiM: a system-level energy evaluation framework for computing-in-memory architectures. arXiv preprint arXiv:1901.09348
Han S, Mao H, Dally WJ (2015) Deep compression: compressing deep neural networks with pruning, trained quantization and Huffman coding. arXiv preprint arXiv:1510.00149
Hsieh K, Khan S, Vijaykumar N, Chang KK, Boroumand A, Ghose S, Mutlu O (2016) Accelerating pointer chasing in 3d-stacked memory: challenges, mechanisms, evaluation. In: 2016 IEEE 34th International conference on computer design (ICCD). IEEE, pp 25–32
Ielmini D (2011) Modeling the universal set/reset characteristics of bipolar RRAM by field-and temperature-driven filament growth. IEEE Trans Electron Devices 58(12):4309–4317
Ielmini D, Nardi F, Cagli C (2010) Resistance-dependent amplitude of random telegraph-signal noise in resistive switching memories. Appl Phys Lett 96(5):053503
Ielmini D, Wong HSP (2018) In-memory computing with resistive switching devices. Nat Electron 1(6):333–343
Jain S, Ranjan A, Roy K, Raghunathan A (2017) Computing in memory with spin-transfer torque magnetic ram. IEEE Trans Very Large Scale Integr (VLSI) Syst 26(3):470–483
Jiang W, Lou Q, Yan Z, Yang L, Hu J, Hu XS, Shi Y (2020) Device-circuit-architecture co-exploration for computing-in-memory neural accelerators. IEEE Trans Comput
Kim KM, Jeong DS, Hwang CS (2011) Nanofilamentary resistive switching in binary oxide system; a review on the present status and outlook. Nanotechnology 22(25):254002
Krizhevsky A et al (2009) Learning multiple layers of features from tiny images
Kvatinsky S, Belousov D, Liman S, Satat G, Wald N, Friedman EG, Kolodny A, Weiser UC (2014) Magic—memristor-aided logic. IEEE Trans Circ Syst II Express Briefs 61(11):895–899
LeCun Y, Bottou L, Bengio Y, Haffner P (1998) Gradient-based learning applied to document recognition. Proc IEEE 86(11):2278–2324
Lee MJ, Lee CB, Lee D, Lee SR, Chang M, Hur JH, Kim YB, Kim CJ, Seo DH, Seo S et al (2011) A fast, high-endurance and scalable non-volatile memory device made from asymmetric Ta2O5−x/TaO2−x bilayer structures. Nat Mater 10(8):625–630
Li KS, Chen PG, Lai TY, Lin CH, Cheng CC, Chen CC, Wei YJ, Hou YF, Liao MH, Lee MH et al (2015) Sub-60 mV-swing negative-capacitance FinFET without hysteresis. In: 2015 IEEE International electron devices meeting (IEDM). IEEE, pp 22–6
Li S, Xu C, Zou Q, Zhao J, Lu Y, Xie Y (2016) Pinatubo: a processing-in-memory architecture for bulk bitwise operations in emerging non-volatile memories. In: Proceedings of the 53rd annual design automation conference, pp 1–6
Li X, Sampson J, Khan A, Ma K, George S, Aziz A, Gupta SK, Salahuddin S, Chang MF, Datta S et al (2017) Enabling energy-efficient nonvolatile computing with negative capacitance FET. IEEE Trans Electron Devices 64(8):3452–3458
Liu Q, Sun J, Lv H, Long S, Yin K, Wan N, Li Y, Sun L, Liu M (2012) Real-time observation on dynamic growth/dissolution of conductive filaments in oxide-electrolyte-based ReRAM. Adv Mater 24(14):1844–1849
Locatelli N, Cros V, Grollier J (2014) Spin-torque building blocks. Nat Mater 13(1):11–20
Loke D, Lee T, Wang W, Shi L, Zhao R, Yeo Y, Chong T, Elliott S (2012) Breaking the speed limits of phase-change memory. Science 336(6088):1566–1569
Mai K, Paaske T, Jayasena N, Ho R., Dally WJ, Horowitz M (2000) Smart memories: a modular reconfigurable architecture. In: Proceedings of 27th international symposium on computer architecture (IEEE Cat. No. RS00201). IEEE, pp 161–171
Niu D, Xiao Y, Xie Y (2012) Low power memristor-based ReRAM design with error correcting code. In: 17th Asia and South Pacific design automation conference. IEEE, pp 79–84
Oskin M, Chong FT, Sherwood T (1998) Active pages: a computation model for intelligent memory. In: Proceedings of the 25th annual international symposium on computer architecture (Cat. No. 98CB36235). IEEE, pp 192–203
Reis D, Niemier M, Hu XS (2018) Computing in memory with FeFETs. In: Proceedings of the international symposium on low power electronics and design, pp 1–6
Shafiee A, Nag A, Muralimanohar N, Balasubramonian R, Strachan JP, Hu M, Williams RS, Srikumar V (2016) Isaac: a convolutional neural network accelerator with in-situ analog arithmetic in crossbars. ACM SIGARCH Comput Architect News 44(3):14–26
Sharma P, Tapily K, Saha A, Zhang J, Shaughnessy A, Aziz A, Snider G, Gupta S, Clark R, Datta S (2017) Impact of total and partial dipole switching on the switching slope of gate-last negative capacitance FETs with ferroelectric hafnium zirconium oxide gate stack. In: 2017 Symposium on VLSI technology. IEEE, pp T154–T155
Slonczewski JC (1996) Current-driven excitation of magnetic multilayers. J Magn Magn Mater 159(1–2):L1–L7
Sze V, Chen YH, Yang TJ, Emer JS (2017) Efficient processing of deep neural networks: a tutorial and survey. Proc IEEE 105(12):2295–2329
Van Lint J, van der Geer G (2012) Introduction to coding theory and algebraic geometry, vol 12. Birkhäuser
Wang D, George S, Aziz A, Datta S, Narayanan V, Gupta SK (2016) Ferroelectric transistor based non-volatile flip-flop. In: Proceedings of the 2016 international symposium on low power electronics and design, pp 10–15
Wang K, Liu Z, Lin Y, Lin J, Han S (2019) HAQ: hardware-aware automated quantization with mixed precision. In: Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp 8612–8620
Xia L, Liu M, Ning X, Chakrabarty K, Wang Y (2017) Fault-tolerant training with on-line fault detection for RRAM-based neural computing systems. In: Proceedings of the 54th annual design automation conference 2017, pp 1–6
Xu C, Niu D, Muralimanohar N, Jouppi NP, Xie Y (2013) Understanding the trade-offs in multi-level cell ReRAM memory design. In: 2013 50th ACM/EDAC/IEEE Design automation conference (DAC). IEEE, pp 1–6
Yan Z, Jiang W, Hu XS, Shi Y (2021) Radars: memory efficient reinforcement learning aided differentiable neural architecture search. arXiv preprint arXiv:2109.05691
Yan Z, Juan DC, Hu XS, Shi Y (2021) Uncertainty modeling of emerging device based computing-in-memory neural accelerators with application to neural architecture search. In: 2021 26th Asia and South Pacific design automation conference (ASP-DAC). IEEE, pp 859–864
Yan Z, Shi Y, Liao W, Hashimoto M, Zhou X, Zhuo C (2020) When single event upset meets deep neural networks: observations, explorations, and remedies. In: 2020 25th Asia and South Pacific design automation conference (ASP-DAC). IEEE, pp 163–168
Yang JJ, Strukov DB, Stewart DR (2013) Memristive devices for computing. Nat Nanotechnol 8(1):13–24
Yang L, Yan Z, Li M, Kwon H, Lai L, Krishna T, Chandra V, Jiang W, Shi Y (2020) Co-exploration of neural architectures and heterogeneous ASIC accelerator designs targeting multiple tasks. In: 2020 57th ACM/IEEE Design automation conference (DAC). IEEE, pp 1–6
Yuasa S, Nagahama T, Fukushima A, Suzuki Y, Ando K (2004) Giant room-temperature magnetoresistance in single-crystal Fe/MgO/Fe magnetic tunnel junctions. Nat Mater 3(12):868–871
Zaman KS, Reaz MBI, Ali SHM, Bakar AAA, Chowdhury MEH (2021) Custom hardware architectures for deep learning on portable devices: a review. IEEE Trans Neural Networks Learn Syst
Zhang D, Jayasena N, Lyashevsky A, Greathouse JL, Xu L, Ignatowski M (2014) TOP-PIM: throughput-oriented programmable processing in memory. In: Proceedings of the 23rd international symposium on high-performance parallel and distributed computing, pp 85–98
Zhao M, Wu H, Gao B, Zhang Q, Wu W, Wang S, Xi Y, Wu D, Deng N, Yu S et al (2017) Investigation of statistical retention of filamentary analog RRAM for neuromorphic computing. In: 2017 IEEE International electron devices meeting (IEDM). IEEE, pp 39–4
Zoph B, Le QV (2017) Neural architecture search with reinforcement learning. In: International conference on learning representations (ICLR)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2023 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this chapter
Cite this chapter
Yan, Z., Hu, X.S., Shi, Y. (2023). On the Reliability of Computing-in-Memory Accelerators for Deep Neural Networks. In: Wang, L., Pattabiraman, K., Di Martino, C., Athreya, A., Bagchi, S. (eds) System Dependability and Analytics. Springer Series in Reliability Engineering. Springer, Cham. https://doi.org/10.1007/978-3-031-02063-6_9
Download citation
DOI: https://doi.org/10.1007/978-3-031-02063-6_9
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-02062-9
Online ISBN: 978-3-031-02063-6
eBook Packages: EngineeringEngineering (R0)