Search

Harnessing multibit precision in nonvolatile memory (NVM)‐based synaptic core can accelerate multiply and accumulate (MAC) operation of deep neural network (DNN). However, NVM‐based synaptic cores suffer from the trade‐off between bit density and performance. The undesired performance degradation with scaling, limited bit precision, and asymmetry associated with weight update poses a severe bottleneck in realizing a high‐density synaptic core. Herein, 1) evaluation of novel differential mode ferroelectric field‐effect transistor (DM‐FeFET) bitcell on a crossbar array of 4 K devices; 2) validation of weighted sum operation on 28 nm DM‐FeFET crossbar array; 3) bit density of 223Mb mm−2, which is ≈2× improvement compared to conventional FeFET array; 4) 196 TOPS/W energy efficiency for VGG‐8 network; and 5) superior bit error rate (BER) resilience showing ≈94% training and 88% inference accuracy with 1% BER are demonstrated. [ABSTRACT FROM AUTHOR]

In this work we report on the three dominant trapping mechanisms affecting the dynamic performance of a double-heterostructure GaN-based MIS-HEMT grown on silicon substrate. In the OFF-state, with high drain voltage and pinched-off 2DEG, the dominant mechanism is the charge-trapping in the gate-drain access region caused by the transversal drain-to-substrate potential. This effect causes the dynamic increase of the ON-resistance, and is positively temperature-dependent, thus of great concern for high-temperature operation. In the SEMI-ON-state, due to the presence of high VDS and relatively high IDS, an additional trapping mechanism emerges, involving the injection of hot electrons from the 2DEG into trap states located in the GaN-buffer or in the AlGaN barrier. This mechanism, critical in hard-switching operations, affects both the ON-resistance and the VTH. Finally, when the gate is positively biased (gate overdrive state) trapping of electrons happens in the gate dielectric layer(s), leading to strong metastable VTH instabilities. [ABSTRACT FROM AUTHOR]

Herein, the gamma‐ray irradiation effect on ferroelectric HfAlOx capacitors for biomedical and space applications is investigated. The experimental results focus on the observation of ferroelectric polarization characteristics in support of different defect mechanisms induced by annealing, endurance cycling, and gamma‐ray irradiation. It is found that the HfAlOx with low Al doping of 6.5% become more vulnerable to radiation exposure, but a remarkable improvement in 9%‐Al‐doped HfAlOx due to less oxygen vacancies is demonstrated. In addition, the results also confirm that the fatigue property of ferroelectric HfAlOx capacitor is determined by the electric‐stress‐induced defect during endurance cycling, not mainly contributed by irradiation damage, which shows the potential for memory applications in rigorous irradiation environment. [ABSTRACT FROM AUTHOR]