IEEE Solid-States Circuits Magazine - Fall 2022 - 27

[14] J.-W. Jang et al., " Sparsity-aware and re
-configurable NPU architecture for Samsung
flagship mobile soc, " in Proc. 2021
ACM/IEEE 48th Annu. Int. Symp. Comput.
Archit.
(ISCA), pp. 15-28, doi: 10.1109/
ISCA52012.2021.00011.
[15] G. Jeong et al., " Union: A unified HW-SW codesign
ecosystem in MLIR for evaluating
tensor operations on spatial accelerators, "
in Proc. 2021 30th Int. Conf. Parallel Archit.
Compilation Techn. (PACT), pp. 30-44, doi:
10.1109/PACT52795.2021.00010.
[16] H. Jia et al., " A programmable neuralnetwork
inference accelerator based on
scalable in-memory computing, " in Proc.
2021 IEEE Int. Solid-State Circuits Conf.
(ISSCC), vol. 64, pp. 236-238, doi: 10.1109/
ISSCC42613.2021.9365788.
[17] N. P. Jouppi et al., " In-datacenter performance
analysis of a tensor processing
unit, " in Proc. 2017 ACM/IEEE 44th Annu.
Int. Symp. Comput. Archit. (ISCA), pp. 1-12,
doi: 10.1145/3079856.3080246.
[18] H. Kwon, L. Lai, M. Pellauer, T. Krishna, Y.H.
Chen, and V. Chandra, " Heterogeneous
dataflow accelerators for multi-DNN workloads, "
in Proc. 2021 IEEE Int. Symp. HighPerform.
Comput. Archit. (HPCA), pp. 71-83,
doi: 10.1109/HPCA51647.2021.00016.
[19] C.-H. Lin et al., " A 3.4-to-13.3TOPS/W
3.6TOPS dual-core deep-learning accelerator
for versatile AI applications in 7nm
5G smartphone SoC, " in Proc. 2020 IEEE
Int. Solid-State Circuits Conf. (ISSCC), pp.
134-136, doi: 10.1109/ISSCC19947.2020.
9063111.
[20] Z.-G. Liu, P. N. Whatmough, and M. Mattina,
" Systolic tensor array: An efficient structured-sparse
GEMM accelerator for mobile
CNN inference, " IEEE Comput. Archit. Lett.,
vol. 19, no. 1, pp. 34-37, Jan./Jun. 2020,
doi: 10.1109/LCA.2020.2979965.
[21] H. Mao et al., " Exploring the granularity
of sparsity in convolutional neural networks, "
in Proc. IEEE Conf. Comput. Vis.
Pattern Recognit. Workshops, 2017, pp.
1927-1934, doi: 10.1109/CVPRW.2017.241.
[22] L. Mei, P. Houshmand, V. Jain, S. Giraldo,
and M. Verhelst, " ZigZag: Enlarging
joint architecture-mapping design space
exploration for DNN accelerators, " IEEE
Trans. Comput., vol. 70, no. 8, pp. 1160-
1174, Aug. 2021, doi: 10.1109/TC.2021.
3059962.
[23] L. Mei, H. Liu, T. Wu, H. E. Sumbul, M. Verhelst,
and E. Beigne, " A uniform latency
model for DNN accelerators with diverse
architectures and dataflows, " in Proc.
2022 Des., Automat. Test Eur. Conf. Exhib.
(DATE), pp. 220-225, doi: 10.23919/
DATE54114.2022.9774728.
[24] B. Moons, R. Uytterhoeven, W. Dehaene, and
M. Verhelst, " Envision: A 0.26-to-10tops/w
subword-parallel dynamic-voltage-accuracyfrequency-scalable
convolutional neural
network processor in 28nm FDSOI, " in
Proc. 2017 IEEE Int. Solid-State Circuits
Conf. (ISSCC), pp. 246-247, doi: 10.1109/
ISSCC.2017.7870353.
[25] " NVIDIA deep learning accelerator (NVDLA). "
NVIDIA. Accessed: 2022. [Online].
Available: http://nvdla.org/
[26] A. Parashar et al., " Timeloop: A systematic
approach to DNN accelerator evaluation, "
in Proc. 2019 IEEE Int. Symp. Perform.
Anal. Syst. Softw. (ISPASS), pp. 304-315,
doi: 10.1109/ISPASS.2019.00042.
[27] N. Shah, L. I. G. Olascoaga, S. Zhao, W. Meert,
and M. Verhelst, " DPU: DAG processing unit
for irregular graphs with precision-scalable
posit arithmetic in 28 nm, " IEEE J. Solid-State
Circuits, vol. 57, no. 8, pp. 2586-2596, Aug.
2022, doi: 10.1109/JSSC.2021.3134897.
[28] Y. S. Shao et al., " Simba: Scaling deeplearning
inference with multi-chip-module-based
architecture, " in Proc. 52nd
Annu. IEEE/ACM Int. Symp. Microarchit.,
ser. MICRO '52. New York, NY, USA:
Association for Computing Machinery,
2019, pp. 14-27, doi: 10.1145/3352460.
3358302.
[29] M. Shi, P. Houshmand, L. Mei, and M.
Verhelst, " Hardware-efficient
residual
neural network execution in line-buffer
depth-first processing, " IEEE J. Emerg.
Sel. Topics Circuits Syst., vol. 11, no. 4, pp.
690-700, Dec. 2021, doi: 10.1109/JETCAS.
2021.3120103.
[30] J. Song et al ., " An 11.5TOPS/W 1024MAC
butterf ly structure dual-core
sparsity-aware neural processing unit
in 8nm flagship mobile SoC, "
2019 IEEE Int. Solid-State Circuits Conf.
(ISSCC), pp. 130-132, doi: 10.1109/ISSCC.
2019.8662476.
[31] A. Symons, L. Mei, and M. Verhelst, " LOMA:
Fast auto-scheduling on DNN accelerators
through loop-order-based memory allocation, "
in Proc. 2021 IEEE 3rd Int. Conf.
Artif. Intell. Circuits Syst. (AICAS), pp. 1-4,
doi: 10.1109/AICAS51828.2021.9458493.
[32] E. Talpes et al., " Compute solution for
Tesla's full self-driving computer, " IEEE
Micro, vol. 40, no. 2, pp. 25-35, Mar./Apr.
2020, doi: 10.1109/MM.2020.2975764.
[33] K. Ueyoshi et al., " DIANA: An end-toend
energy-efficient digital and ANAlog
hybrid neural network SoC, " in Proc.
2022 IEEE Int. Solid-State Circuits Conf.
(ISSCC), vol. 65, pp. 1-3, doi: 10.1109/
ISSCC42614.2022.9731716.
[34] R. Venkatesan et al., " MAGNet: A modular
accelerator generator for neural networks, "
in Proc. 2019 IEEE/ACM Int. Conf.
Comput.-Aided Des. (ICCAD), pp. 1-8, doi:
10.1109/ICCAD45719.2019.8942127.
[35] W. Wan et al., " A compute-in-memory
chip based on resistive random-access
memory, " Nature, vol. 608, no. 7923, pp.
504-512, Aug. 2022, doi: 10.1038/s41586
-022-04992-8.
[36] X. Yang et al., " Interstellar: Using Halide's
scheduling language to analyze DNN accelerators, "
in Proc. 25th Int. Conf. Archit.
Support Program. Lang. Oper. Syst., ser.
ASPLOS' 20. New York, NY, USA: Association
for Computing Machinery, 2020, pp.
369-383, doi: 10.1145/3373376.3378514.
[37] Z. Zhao, H. Kwon, S. Kuhar, W. Sheng, Z.
Mao, and T. Krishna, " mRNA: Enabling
efficient mapping space exploration for
a reconfiguration neural accelerator, " in
Proc. 2019 IEEE Int. Symp. Perform. Anal.
Syst. Softw. (ISPASS), pp. 282-292, doi:
10.1109/ISPASS.2019.00040.
About the Authors
Marian Verhelst (marian.verhelst@
kuleuven.be) received her Ph.D. degree
from KU Leuven in 2008 and
worked as a research scientist at Intel
Labs, Hillsboro, OR, USA. She is a full
professor at the MICAS Laboratories,
Electrical Engineering Department
(ESAT), KU Leuven 3001, Belgium, and
a research director at imec. Her research
focuses on embedded machine
learning, hardware accelerators, HWalgorithm
codesign, and low-power
in Proc.
edge processing. She is active in the
TPCs of DATE, ISSCC, VLSI, and ESSCIRC
and was the chair of tinyML2021
and the TPC cochair of AICAS2020.
She is an IEEE SSCS Distinguished
Lecturer; was a member of the Young
Academy of Belgium; was an associate
editor for IEEE Transactions on
Very Large Scale Integration (VLSI)
Systems, IEEE Transactions on Circuits
and Systems II: Express Briefs, and IEEE
Journal of Solid-State Circuits; and is
a member of the STEM advisory committee
to the Flemish Government.
She currently holds a prestigious ERC
Starting Grant from the European
Union; was the laureate of the Royal
Academy of Belgium in 2016; and received
the André Mischke YAE Prize
for Science and Policy in 2021.
Man Shi (man.shi@kuleuven.be)
received her B.Sc. degree from the
School of Information Science and Engineering
from Shandong University,
China, in 2017 and her M.Sc. degree
from the Institute of Microelectronics,
Tsinghua University, China, in 2020.
She is currently pursuing a Ph.D. degree
in accelerators architecture for
deep neural networks with the MICAS
Laboratories, Electrical Engineering
Department (ESAT), KU Leuven 3000,
Belgium. Her current research interests
include low-power deep neural
network hardware accelerator design;
algorithm-hardware codesign; and
reconfigured computation.
Linyan Mei (linyan.mei@kuleuven.
be) received her B.Sc. degree in electronic
science and technology from
the Beijing Institute of Technology,
China, in 2016, and her M.Sc. in electrical
engineering from KU Leuven,
Belgium, in 2018. She is currently pursuing
a Ph.D. in digital design for embedded
machine learning processors
with the MICAS Laboratories, Electrical
Engineering Department (ESAT), KU
Leuven 3000, Belgium. She was an intern
with imec, Leuven, Belgium, from
2017 to 2018 and an intern with Meta,
Menlo Park, CA, USA, in the winter of
2021. Her current research interests
include design space exploration for
deep neural network accelerators and
precision-scalable computing.
IEEE SOLID-STATE CIRCUITS MAGAZINE
FALL 2022
27
http://www.nvdla.org/
IEEE Solid-States Circuits Magazine - Fall 2022

Table of Contents for the Digital Edition of IEEE Solid-States Circuits Magazine - Fall 2022
