A Quantization Model Based on a Floating-point Computing-in-Memory Architecture | IEEE Conference Publication | IEEE Xplore