We present a 256 × 256 in-memory compute (IMC) core designed and fabricated in 14-nm CMOS technology with backend-integrated multi-level phase change memory (PCM). It comprises 256 linearized current-controlled oscillator (CCO)-based A/D converters (ADCs) at a compact 4-μm pitch and a local digital processing unit (LDPU) performing affine scaling and ReLU operations. A frequency-linearization technique for CCO is introduced, which increases the maximum CCO frequency beyond 3 GHz, while ensuring accurate on-chip matrix-vector multiplications (MVMs). Moreover, the design and functionality of the digital ADC calibration procedure is described in detail and the MVM accuracy is quantified. Finally, the measured classification accuracies of deep learning (DL) inference applications on the MNIST and CIFAR-10 datasets, when two IMC cores are employed, are presented. For a performance density of 1.59 TOPS/mm2, a measured energy efficiency of 10.5 TOPS/W, at a main clock frequency of 1 GHz, is achieved.

@en

As a tide propagates into the estuary, river discharge affects tidal damping, primarily via a friction term, attenuating tidal motion by increasing the quadratic velocity in the numerator, while reducing the effective friction by increasing the water depth in the denominator. For the first time, we demonstrate a third effect of river discharge that may lead to the weakening of the channel convergence (i.e. landward reduction of channel width and/or depth). In this study, monthly averaged tidal water levels (2003-2014) at six gauging stations along the Yangtze River estuary are used to understand the seasonal behaviour of tidal damping and residual water level slope. Observations show that there is a critical value of river discharge, beyond which the tidal damping is reduced with increasing river discharge. This phenomenon is clearly observed in the upstream part of the Yangtze River estuary (between the Maanshan and Wuhu reaches), which suggests an important cumulative effect of residual water level on tide-river dynamics. To understand the underlying mechanism, an analytical model has been used to quantify the seasonal behaviour of tide-river dynamics and the corresponding residual water level slope under various external forcing conditions. It is shown that a critical position along the estuary is where there is maximum tidal damping (approximately corresponding to a maximum residual water level slope), upstream of which tidal damping is reduced in the landward direction. Moreover, contrary to the common assumption that larger river discharge leads to heavier damping, we demonstrate that beyond a critical value tidal damping is slightly reduced with increasing river discharge, owing to the cumulative effect of the residual water level on the effective friction and channel convergence. Our contribution describes the seasonal patterns of tide-river dynamics in detail, which will, hopefully, enhance our understanding of the nonlinear tide-river interplay and guide effective and sustainable water management in the Yangtze River estuary and other estuaries with substantial freshwater discharge.@en