23 November 2022 in IEEE Transactions on Circuits and Systems II: Express Briefs
This brief introduces a machine learning based framework to model FinFET’s I-V and C-V curves with artificial neural networks and to further optimize FinFET’s performance on DC and AC characteristics. Our ML-based device model for FinFETs takes $V_{gs}$ and other nine parameters that define geometry, doping, stress, and work function profile as input variables. Experimental results show that our ML-based device model can predict current and capacitance accurately. Besides, the proposed ML-based performance optimization flow is a promising alternative to the traditional design of experiment method based on technology computer-aided design simulations. Our optimization flow can locate optimized device features accurately and require fewer simulations. Our work demonstrates that ML can perform compact modeling of advanced devices with high accuracy, and the trained ANN models can accelerate performance optimization.
27 June 2024 in IEEE Transactions on Circuits and Systems for Video Technology
Neural Volume Rendering (NVR) has advanced explosively since the advent of Neural Radiance Field (NeRF), a technique for novel view synthesis of complex scenes based on a finite set of input views. Existing ray casting-based NVR approaches process rays concurrently to leverage parallelism but fails to consider its impact on cache locality, which ultimately undermines the efficiency of corresponding dedicated hardware accelerator designs. We further observed that there exhibits spatial correspondence between features and voxels in NVR that can be exploited by processing in the order of voxel, not ray. This paper introduces a novel approach to meticulously reorder the execution of rays, ensuring that rays with similar memory access patterns are processed in parallel, thereby enhancing cache locality. On the basis of that, we also propose an efficient backend architecture and a corresponding memory subsystem, facilitating accurate data prefetching to hide off-chip memory latency. To validate the proposed architecture, we implement our design in VerilogHDL and evaluate the performance by post-synthesis simulation with real scene data. The evaluation results demonstrate that our design markedly enhances the efficiency of NVR processing, achieving a considerable speedup (1.62×) compared to the state-of-the-art NVR accelerator, while necessitating significantly less silicon area (5.12×) and power (32.79×).
30 December 2022 in IEEE Transactions on Circuits and Systems I: Regular Papers
This work presents an energy-efficient accelerator for medical image reconstruction from implicit neural representation (INR). The accelerator implements an INR-based algorithm to deliver high-quality medical image reconstruction with arbitrary resolution from a compact implicit format. In particular, we propose a dedicated hardware architecture based on an optimized computation flow for the INR-based reconstruction algorithm, which co-designs data reuse and computation load. The proposed architecture takes in the coordinate of the intersection of three scans and outputs all the voxel intensities, minimizing the data movement between on-chip and off-chip. To validate the proposed accelerator, we build a proof-of-concept prototype demonstration system using field programmable gate array (FPGA). We also map our design to 40nm CMOS technology to measure the performance of the proposed accelerator. The implementation results show that, running at 400MHz, the proposed accelerator is capable of processing medical images with $256\times 256$ resolution in real-time at 26.3 frames per second (FPS), with a power consumption of only 795 mW. Comparison results show that the performance, as well as the energy efficiency of the proposed accelerator, outperforms the central processing unit (CPU)-based and graphic processing unit (GPU)-based implementations.
17 March 2023 in IEEE Transactions on Circuits and Systems II: Express Briefs
Latency is one of the critical performance metrics for Networks-on-Chips (NoCs). When designing an NoC, the designers have to explore enormous design parameters and various traffic patterns, thus a fast and accurate latency estimator is essential to explore the large design space. In this paper, we present an ideal neural network-based methodology for latency estimation in NoCs, especially for application-specific traffics. By inputting the sequence of extracted traffic features and NoC parameters, the neural network model will infer the corresponding average network latency in a fast while accurate way. Instead of training one neural network model for each benchmark from scratch, we adopt transfer learning to train the network model for a new benchmark from another trained one. Experimental results on a set of widely used application-specific NoC benchmarks show that, our method can achieve an average estimation accuracy of 95%, and a 17.1X speedup for large NoCs compared to BookSim2 simulations. Our method can also achieve 20% to 70% improvement in accuracy over the other state-of-art machine learning-based works.
10 August 2021 in IEEE Transactions on Electron Devices
FinFET architecture has witnessed great success at commercial 22-nm node and beyond due to its natural immunity of short-channel effects (SCEs) since the past decade. As the critical size scales down further, the parasitic capacitance becomes a crucial bottleneck to gain the device performance. In this article, analytical parasitic capacitance models (including eight components) are established based on real Si-bulk FinFET pMOS. Compared with the prior works, first, the electric field distribution with a small gate voltage perturbation is employed to reveal the capacitance origin and distribution. Second, except for widely studied parameters (spacer thickness and overlap), the key parameters [source and drain (S/D) recess depth, spacer height, and embedded-SiGe (e-SiGe) cavity overfill] are also studied. The later parameters are seldom considered, but they are widely selected as key in-line monitor items in manufacturing due to their sensitivity to device current and parasitic capacitance. In addition, the geometric dependence on parasitic capacitance is performed and verified well by the 3-D Sentaurus Technology Computer Aided Design (TCAD) simulator. The result shows that the analytical capacitance model can capture geometric behavior well, which finally is understood by the breakdown of the total parasitic capacitance.
31 July 2024 in ACM Journal on Emerging Technologies in Computing Systems
Artificial Neural Networks (ANNs) have achieved remarkable performance in many artificial intelligence tasks. As the application scenarios become more sophisticated, the computation and energy consumption of ANNs are also constantly increasing, which poses a challenge for deploying ANNs on energy-constrained devices. Spiking Neural Networks (SNNs) provide a promising solution to build energy-efficiency neural networks. However, the current training methods of SNNs cannot output values as precise as ANNs. This limits the applications of SNNs to relatively simple image classification tasks. In this article, we extend the application of SNNs to neural rendering tasks and propose an energy-efficient spiking neural rendering model, called Spiking-NeRF (Spiking Neural Radiance Fields). We first analyze the ANN-to-SNN conversion theory and propose an output scheme for SNNs to obtain the precise scene property values. Then we customize the parameter normalization method for the special network architecture of neural rendering. Furthermore, we present an early termination strategy (ETS) based on the discrete nature of spikes to reduce energy consumption. We evaluate the performance of Spiking-NeRF on both realistic and synthetic scenes. Experimental results show that Spiking-NeRF can achieve comparable rendering performance to ANN-based NeRF with up to \(2.27\times\) energy reduction.