Xuehai Zhou

Bio: Xuehai Zhou is an academic researcher from University of Science and Technology of China. The author has contributed to research in topics: Speedup & Field-programmable gate array. The author has an hindex of 24, co-authored 244 publications receiving 2447 citations.

PuDianNao: A Polyvalent Machine Learning Accelerator

Abstract: Machine Learning (ML) techniques are pervasive tools in various emerging commercial applications, but have to be accommodated by powerful computer systems to process very large data. Although general-purpose CPUs and GPUs have provided straightforward solutions, their energy-efficiencies are limited due to their excessive supports for flexibility. Hardware accelerators may achieve better energy-efficiencies, but each accelerator often accommodates only a single ML technique (family). According to the famous No-Free-Lunch theorem in the ML domain, however, an ML technique performs well on a dataset may perform poorly on another dataset, which implies that such accelerator may sometimes lead to poor learning accuracy. Even if regardless of the learning accuracy, such accelerator can still become inapplicable simply because the concrete ML task is altered, or the user chooses another ML technique. In this study, we present an ML accelerator called PuDianNao, which accommodates seven representative ML techniques, including k-means, k-nearest neighbors, naive bayes, support vector machine, linear regression, classification tree, and deep neural network. Benefited from our thorough analysis on computational primitives and locality properties of different ML techniques, PuDianNao can perform up to 1056 GOP/s (e.g., additions and multiplications) in an area of 3.51 mm^2, and consumes 596 mW only. Compared with the NVIDIA K20M GPU (28nm process), PuDianNao (65nm process) is 1.20x faster, and can reduce the energy by 128.41x.

DLAU: A Scalable Deep Learning Accelerator Unit on FPGA

Abstract: As the emerging field of machine learning, deep learning shows excellent ability in solving complex learning problems. However, the size of the networks becomes increasingly large scale due to the demands of the practical applications, which poses significant challenge to construct a high performance implementations of deep learning neural networks. In order to improve the performance as well as to maintain the low power cost, in this paper we design deep learning accelerator unit (DLAU), which is a scalable accelerator architecture for large-scale deep learning networks using field-programmable gate array (FPGA) as the hardware prototype. The DLAU accelerator employs three pipelined processing units to improve the throughput and utilizes tile techniques to explore locality for deep learning applications. Experimental results on the state-of-the-art Xilinx FPGA board demonstrate that the DLAU accelerator is able to achieve up to $36.1 {\times }$ speedup comparing to the Intel Core2 processors, with the power consumption at 234 mW.

Cambricon-s: addressing irregularity in sparse neural networks through a cooperative software/hardware approach

Abstract: Neural networks have become the dominant algorithms rapidly as they achieve state-of-the-art performance in a broad range of applications such as image recognition, speech recognition and natural language processing. However, neural networks keep moving towards deeper and larger architectures, posing a great challenge to the huge amount of data and computations. Although sparsity has emerged as an effective solution for reducing the intensity of computation and memory accesses directly, irregularity caused by sparsity (including sparse synapses and neurons) prevents accelerators from completely leveraging the benefits; it also introduces costly indexing module in accelerators. In this paper, we propose a cooperative software/hardware approach to address the irregularity of sparse neural networks efficiently. Initially, we observe the local convergence, namely larger weights tend to gather into small clusters during training. Based on that key observation, we propose a software-based coarse-grained pruning technique to reduce the irregularity of sparse synapses drastically. The coarse-grained pruning technique, together with local quantization, significantly reduces the size of indexes and improves the network compression ratio. We further design a hardware accelerator, Cambricon-S, to address the remaining irregularity of sparse synapses and neurons efficiently. The novel accelerator features a selector module to filter unnecessary synapses and neurons. Compared with a state-of-the-art sparse neural network accelerator, our accelerator is 1.71× and 1.37× better in terms of performance and energy efficiency, respectively.

MALOC: A Fully Pipelined FPGA Accelerator for Convolutional Neural Networks With All Layers Mapped on Chip

Abstract: Recently, field-programmable gate arrays (FPGAs) have been widely used in the implementations of hardware accelerator for convolutional neural networks (CNNs). However, most of these existing accelerators are designed in the same idea as their ASIC counterparts, in which all operations from different layers are mapped to the same hardware units and working in a multiplexed way. This manner does not take full advantage of reconfigurability and customizability of FPGAs, resulting in a certain degree of computational efficiency degradation. In this paper, we propose a new architecture for FPGA-based CNN accelerator that maps all the layers to their own on-chip units and working concurrently as a pipeline. A comprehensive mapping and optimizing methodology based on establishing roofline model oriented optimization model is proposed, which can achieve maximum resource utilization as well as optimal computational efficiency. Besides, to ease the programming burden, we propose a design framework which can provide a one-stop function for developers to generate the accelerator with our optimizing methodology. We evaluate our proposal by implementing different modern CNN models on Xilinx Zynq-7020 and Virtex-7 690t FPGA platforms. Experimental results show that our implementations can achieve a peak performance of 910.2 GOPS on Virtex-7 690t, and 36.36 GOP/s/W energy efficiency on Zynq-7020, which are superior to the previous approaches.

WGAN-Based Synthetic Minority Over-Sampling Technique: Improving Semantic Fine-Grained Classification for Lung Nodules in CT Images

Abstract: Data imbalance issue generally exists in most medical image analysis problems and maybe getting important with the popularization of data-hungry deep learning paradigms. We explore the cutting-edge Wasserstein generative adversarial networks (WGANs) to address the data imbalance problem with oversampling on the minority classes. The WGAN can estimate the underlying distribution of a minority class to synthesize more plausible and helpful samples for the classification model. In this paper, the WGAN-based over-sampling technique is applied to augment the data to balance for the fine-grained classification of seven semantic attributes of lung nodules in computed tomography images. The fine-grained classification is carried out with a normal convolutional neural network (CNN). To further illustrate the efficacy of the WGAN-based over-sampling technique, the conventional data augmentation method commonly used in many deep learning works, the generative adversarial networks (GANs), and the deep convolutional generative adversarial networks (DCGANs) are implemented for comparison. The whole schemes of the minority oversampling and fine-grained classification are tested with the public lung imaging database consortium dataset. The experimental results suggest that the WGAN-based oversampling technique can synthesize helpful samples for the minority classes to assist the training of the CNN model and to boost the fine-grained classification performance better than the conventional data augmentation method and the two schemes of the GAN and DCGAN techniques do. It may thus suggest that the WGAN technique offers an alternative methodological option for the further deep learning on imbalanced classification studies.

In-Datacenter Performance Analysis of a Tensor Processing Unit

Abstract: Many architects believe that major improvements in cost-energy-performance must now come from domain-specific hardware. This paper evaluates a custom ASIC---called a Tensor Processing Unit (TPU)---deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN). The heart of the TPU is a 65,536 8-bit MAC matrix multiply unit that offers a peak throughput of 92 TeraOps/second (TOPS) and a large (28 MiB) software-managed on-chip memory. The TPU's deterministic execution model is a better match to the 99th-percentile response-time requirement of our NN applications than are the time-varying optimizations of CPUs and GPUs (caches, out-of-order execution, multithreading, multiprocessing, prefetching, ...) that help average throughput more than guaranteed latency. The lack of such features helps explain why, despite having myriad MACs and a big memory, the TPU is relatively small and low power. We compare the TPU to a server-class Intel Haswell CPU and an Nvidia K80 GPU, which are contemporaries deployed in the same datacenters. Our workload, written in the high-level TensorFlow framework, uses production NN applications (MLPs, CNNs, and LSTMs) that represent 95% of our datacenters' NN inference demand. Despite low utilization for some applications, the TPU is on average about 15X - 30X faster than its contemporary GPU or CPU, with TOPS/Watt about 30X - 80X higher. Moreover, using the GPU's GDDR5 memory in the TPU would triple achieved TOPS and raise TOPS/Watt to nearly 70X the GPU and 200X the CPU.

ISAAC: a convolutional neural network accelerator with in-situ analog arithmetic in crossbars

Abstract: A number of recent efforts have attempted to design accelerators for popular machine learning algorithms, such as those involving convolutional and deep neural networks (CNNs and DNNs). These algorithms typically involve a large number of multiply-accumulate (dot-product) operations. A recent project, DaDianNao, adopts a near data processing approach, where a specialized neural functional unit performs all the digital arithmetic operations and receives input weights from adjacent eDRAM banks.This work explores an in-situ processing approach, where memristor crossbar arrays not only store input weights, but are also used to perform dot-product operations in an analog manner. While the use of crossbar memory as an analog dot-product engine is well known, no prior work has designed or characterized a full-fledged accelerator based on crossbars. In particular, our work makes the following contributions: (i) We design a pipelined architecture, with some crossbars dedicated for each neural network layer, and eDRAM buffers that aggregate data between pipeline stages. (ii) We define new data encoding techniques that are amenable to analog computations and that can reduce the high overheads of analog-to-digital conversion (ADC). (iii) We define the many supporting digital components required in an analog CNN accelerator and carry out a design space exploration to identify the best balance of memristor storage/compute, ADCs, and eDRAM storage on a chip. On a suite of CNN and DNN workloads, the proposed ISAAC architecture yields improvements of 14.8×, 5.5×, and 7.5× in throughput, energy, and computational density (respectively), relative to the state-of-the-art DaDianNao architecture.

Critical analysis of Big Data challenges and analytical methods

Abstract: Big Data (BD), with their potential to ascertain valued insights for enhanced decision-making process, have recently attracted substantial interest from both academics and practitioners. Big Data Analytics (BDA) is increasingly becoming a trending practice that many organizations are adopting with the purpose of constructing valuable information from BD. The analytics process, including the deployment and use of BDA tools, is seen by organizations as a tool to improve operational efficiency though it has strategic potential, drive new revenue streams and gain competitive advantages over business rivals. However, there are different types of analytic applications to consider. Therefore, prior to hasty use and buying costly BD tools, there is a need for organizations to first understand the BDA landscape. Given the significant nature of the BD and BDA, this paper presents a state-of-the-art review that presents a holistic view of the BD challenges and BDA methods theorized/proposed/employed by organizations to help others understand this landscape with the objective of making robust investment decisions. In doing so, systematically analysing and synthesizing the extant research published on BD and BDA area. More specifically, the authors seek to answer the following two principal questions: Q1 – What are the different types of BD challenges theorized/proposed/confronted by organizations? and Q2 – What are the different types of BDA methods theorized/proposed/employed to overcome BD challenges? . This systematic literature review (SLR) is carried out through observing and understanding the past trends and extant patterns/themes in the BDA research area, evaluating contributions, summarizing knowledge, thereby identifying limitations, implications and potential further research avenues to support the academic community in exploring research themes/patterns. Thus, to trace the implementation of BD strategies, a profiling method is employed to analyze articles (published in English-speaking peer-reviewed journals between 1996 and 2015) extracted from the Scopus database. The analysis presented in this paper has identified relevant BD research studies that have contributed both conceptually and empirically to the expansion and accrual of intellectual wealth to the BDA in technology and organizational resource management discipline.

PRIME: a novel processing-in-memory architecture for neural network computation in ReRAM-based main memory

Abstract: Processing-in-memory (PIM) is a promising solution to address the "memory wall" challenges for future computer systems. Prior proposed PIM architectures put additional computation logic in or near memory. The emerging metal-oxide resistive random access memory (ReRAM) has showed its potential to be used for main memory. Moreover, with its crossbar array structure, ReRAM can perform matrix-vector multiplication efficiently, and has been widely studied to accelerate neural network (NN) applications. In this work, we propose a novel PIM architecture, called PRIME, to accelerate NN applications in ReRAM based main memory. In PRIME, a portion of ReRAM crossbar arrays can be configured as accelerators for NN applications or as normal memory for a larger memory space. We provide microarchitecture and circuit designs to enable the morphable functions with an insignificant area overhead. We also design a software/hardware interface for software developers to implement various NNs on PRIME. Benefiting from both the PIM architecture and the efficiency of using ReRAM for NN computation, PRIME distinguishes itself from prior work on NN acceleration, with significant performance improvement and energy saving. Our experimental results show that, compared with a state-of-the-art neural processing unit design, PRIME improves the performance by ~2360× and the energy consumption by ~895×, across the evaluated machine learning benchmarks.

Going Deeper with Embedded FPGA Platform for Convolutional Neural Network

Abstract: In recent years, convolutional neural network (CNN) based methods have achieved great success in a large number of applications and have been among the most powerful and widely used techniques in computer vision. However, CNN-based methods are com-putational-intensive and resource-consuming, and thus are hard to be integrated into embedded systems such as smart phones, smart glasses, and robots. FPGA is one of the most promising platforms for accelerating CNN, but the limited bandwidth and on-chip memory size limit the performance of FPGA accelerator for CNN.In this paper, we go deeper with the embedded FPGA platform on accelerating CNNs and propose a CNN accelerator design on embedded FPGA for Image-Net large-scale image classification. We first present an in-depth analysis of state-of-the-art CNN models and show that Convolutional layers are computational-centric and Fully-Connected layers are memory-centric.Then the dynamic-precision data quantization method and a convolver design that is efficient for all layer types in CNN are proposed to improve the bandwidth and resource utilization. Results show that only 0.4% accuracy loss is introduced by our data quantization flow for the very deep VGG16 model when 8/4-bit quantization is used. A data arrangement method is proposed to further ensure a high utilization of the external memory bandwidth. Finally, a state-of-the-art CNN, VGG16-SVD, is implemented on an embedded FPGA platform as a case study. VGG16-SVD is the largest and most accurate network that has been implemented on FPGA end-to-end so far. The system on Xilinx Zynq ZC706 board achieves a frame rate at 4.45 fps with the top-5 accuracy of 86.66% using 16-bit quantization. The average performance of convolutional layers and the full CNN is 187.8 GOP/s and 137.0 GOP/s under 150MHz working frequency, which outperform previous approaches significantly.