Table 4 Summary of reported simulation frameworks for the study of memristive hardware neural networks

From: Hardware implementation of memristor-based artificial neural networks

Simulation Framework	Year	Platform	Training	Simulation type	Open Source	Type of ANN	Compatible dev.	Energy	Accuracy	Power	Latency	Variability	R_L	C_L	CMOS	GPU
Tensorflow²²⁴	2015	Python	Yes	Neural network	Yes	MLP, CNN	No dev.	No	Yes	No	No	Yes	No	No	No	Yes
Pytorch²²³	2017	Python	Yes	Neural network	Yes	MLP, CNN	No dev.	No	Yes	No	No	Yes	No	No	No	Yes
NEURON²⁸⁵	2006	Python	Yes	Neural network	Yes	SNN	No dev.	No	Yes	No	No	Yes	No	No	No	Yes
Brian2²²⁵	2019	Python	Yes	Neural network	Yes	SNN	No dev.	No	Yes	No	No	Yes	No	No	No	Yes
NEST²²⁶	2007	Python	Yes	Neural network	Yes	SNN	No dev.	No	Yes	No	No	Yes	No	No	No	Yes
BindsNET²²⁷	2018	Python	Yes	Neural network	Yes	SNN	No dev.	No	Yes	No	No	Yes	No	No	No	Yes
Memtorch²⁸⁶	2020	Python, C++, CUDA	No	Neurla network	Yes	CNN	RRAM	No	Yes	No	No	Yes	No	No	No	Yes
NVMain^238,239	2015	C++	No	Architecture	Yes	Memory	RRAM	Yes	No	Yes	Yes	No	No	No	No	No
PUMA¹⁵⁴	2019	C++	No	Architecture	No	MLP, CNN	RRAM	Yes	Yes	Yes	Yes	No	No	No	Yes	Yes
RAPIDNN²⁴⁷	2018	C++	No	Architecture	No	MLP, CNN	RRAM	Yes	Yes	Yes	Yes	Yes	No	No	Yes	No
DL-RSIM²²⁸	2018	Python	No	Architecture	No	MLP, CNN	RRAM	No	Yes	No	No	Yes	No	No	No	Yes
PipeLayer²⁴⁶	2017	C++	Yes	Architecture	No	CNN	RRAM	Yes		Yes	Yes	No	No	No	No	No
Tiny but Accurate²³⁰	2019	MATLAB	No	Architecture	Yes	CNN, ResNET	RRAM	Yes	Yes	Yes	No	No	No	No	No	No
Yuan et al.²³¹	2019	C++, MATLAB	Yes	Architecture	Yes	No data	RRAM	No	Yes	Yes	No	No	No	No	No	Yes
Sun et al.²²⁹	2019	Python	Yes	Architecture	No	MLP	PCM, STT-RAM, ReRAM, SRAM, FeFET	Yes	Yes	Yes	No	Yes	No	No	No	No
A. Chen²⁴⁸	2013	MATLAB	No	Circuit	Yes	MLP	RRAM	No	No	Yes	No	Yes	Yes	No	No	No
CIM-SIM²⁴²	2019	SystemC (C++)	No	Architecture	Yes	SLP	RRAM	No	No	No	No	No	No	No	No	No
MNSIM^240,241	2018	Python	No	Architecture	Yes	CNN	RRAM	Yes	No	Yes	Yes	Yes	No	No	No	Yes
NVSIM²⁴⁵	2012	C++	No	Circuital	Yes	Memory	PCM, STT-RAM, ReRAM, Flash	Yes	No	Yes	Yes	No	No	No	No	No
CrossSIM²⁸⁷	2017	Python	No data	Circuital	Yes	No data	PCM, ReRAM, Flash	No	Yes	No	No	Yes	Yes	No	No	Yes
NeuroSIM⁹¹	2022	Python, C++	Yes	Circuital	Yes	MLP, CNN	PCM, STT-RAM, ReRAM, SRAM, FeFET	Yes	Yes	Yes	No	Yes	No	No	No	Yes
NVM-SPICE²⁵²	2012	Not specified	No	Circuital	No	SLP	RRAM	Yes	No	Yes	Yes	Yes	No	No	Yes	No
IBM Analog Hardware Acceleration Kit²³²	2021	Python, C++, CUDA	Yes	Neural network	Yes	MLP, CNN, LSTM	PCM	No	Yes	No	No	Yes	No	No	No	Yes
Fritscher et al.²⁵⁶	2019	Mixed (VHDL, Verilog, SPICE)	No	Circuital	No	MLP	PCM, STT-RAM, ReRAM, SRAM, FeFET	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Aguirre et al.²⁵³	2020	Mixed (Python, MATLAB, SPICE)	No	Circuital	No	MLP	PCM, STT-RAM, ReRAM, SRAM, FeFET	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes

Back to article page

Table 4 Summary of reported simulation frameworks for the study of memristive hardware neural networks

Search

Quick links