SCIENCE CHINA Information Sciences, Volume 64 , Issue 6 : 169401(2021) https://doi.org/10.1007/s11432-020-2942-2

Flash memory based computing-in-memory system to solve partial differential equations

More info
  • ReceivedMar 13, 2020
  • AcceptedJun 4, 2020
  • PublishedSep 23, 2020


There is no abstract available for this article.


This work was supported by China Key Research and Development Program (Grant No. 2016YFA0201802) and National Natural Science Foundation of China (Grant Nos. 91964105, 61874068).


Appendixes A–D.


[1] Ma W, Zidan M A, Lu W D. Neuromorphic computing with memristive devices. Sci China Inf Sci, 2018, 61: 060422 CrossRef Google Scholar

[2] Raoux S, We?nic W, Ielmini D. Phase Change Materials and Their Application to Nonvolatile Memories. Chem Rev, 2010, 110: 240-267 CrossRef Google Scholar

[3] Mikolajick T, Dehm C, Hartner W. FeRAM technology for high density applications. MicroElectron Reliability, 2001, 41: 947-950 CrossRef Google Scholar

[4] Guo X, Bayat F M, Bavandpour M, et al. Fast, energy-efficient, robust, and reproducible mixed-signal neuromorphic classifier based on embedded NOR flash memory technology. In: Proceedings of 2017 IEEE International Electron Devices Meeting (IEDM), 2017. 1--4. Google Scholar

[5] Bavandpour M, Mahmoodi M R, Strukov D B. Energy-Efficient Time-Domain Vector-by-Matrix Multiplier for Neurocomputing and Beyond. IEEE Trans Circuits Syst II, 2019, 66: 1512-1516 CrossRef Google Scholar

[6] Sun Z, Pedretti G, Ambrosi E. Solving matrix equations in one step with cross-point resistive arrays. Proc Natl Acad Sci USA, 2019, 116: 4123-4128 CrossRef ADS Google Scholar

[7] Zidan M A, Jeong Y J, Lee J. A general memristor-based partial differential equation solver. Nat Electron, 2018, 1: 411-420 CrossRef Google Scholar

[8] Siau C, Kim K H, Lee S, et al. 13.5 A 512Gb 3-bit/Cell 3D Flash Memory on 128-Wordline-Layer with 132MB/s Write Performance Featuring Circuit-Under-Array Technology. In: Proceedings of 2019 IEEE International Solid-State Circuits Conference-(ISSCC), 2019. 218--220. Google Scholar

  • Figure 1

    (Color online) (a) Schematic diagram of flash memory cell arrays. The values of the elements are mapped as pulse time and threshold voltage, and the vector-matrix multiplication operations of the slice are performed by supplying the input vector as voltage pulses to the rows and reading out the charge outputs at each string. (b) The way we divide the blocks leads to totally six different slice patterns by enlarging the selected area of 144$\times$144 coefficient matrix generated by the FDM method. (c) Measured output obtained from 144$\times$1 matrix after 50 iterations.