A5100643410- Advanced Data Storage Technologies
- Semiconductor materials and devices
- Parallel Computing and Optimization Techniques
- Radiation Effects in Electronics
- Advancements in Semiconductor Devices and Circuit Design
- Algorithms and Data Compression
- Advanced Data Compression Techniques
- Distributed and Parallel Computing Systems
- Integrated Circuits and Semiconductor Failure Analysis
- Distributed systems and fault tolerance
- Ferroelectric and Negative Capacitance Devices
- Advanced Memory and Neural Computing
- Low-power high-performance VLSI design
- Advanced Neural Network Applications
- Image and Video Quality Assessment
- Image and Signal Denoising Methods
- Advanced Graph Neural Networks
- Caching and Content Delivery
- VLSI and Analog Circuit Testing
- Advanced Image Fusion Techniques
- Interconnection Networks and Systems
- Domain Adaptation and Few-Shot Learning
- Advanced Image Processing Techniques
- Image Enhancement Techniques
- Silicon Carbide Semiconductor Technologies
Florida State University
2023-2025
Taizhou University
2025
Nanjing Tech University
2024
University of Electronic Science and Technology of China
2019-2024
State Key Laboratory of Electronic Thin Films and Integrated Devices
2024
National Engineering Research Center of Electromagnetic Radiation Control Materials
2023-2024
Beijing Polytechnic
2021-2024
Shanghai Aerospace Automobile Electromechanical (China)
2024
Xi'an Shiyou University
2024
Chinese Academy of Sciences
2003-2023
We present a fully integrated 14nm CMOS technology featuring finFET architecture on an SOI substrate for diverse set of SoC applications including HP server microprocessors and LP ASICs. This is with 4 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sup> generation deep trench embedded DRAM to provide ultra-dense (0.0174um xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> ) memory solution industry leading 'scale-out' processor design. A...
Today's scientific simulations are producing vast volumes of data that cannot be stored and transferred efficiently because limited storage capacity, parallel I/O bandwidth, network bandwidth. The situation is getting worse over time the ever-increasing gap between relatively slow transfer speed fast-growing computation power in modern supercomputers. Error-bounded lossy compression becoming one most critical techniques for resolving big issue, it can significantly reduce volume while...
Today's scientific simulations require a significant reduction of data volume because extremely large amounts they produce and the limited I/O bandwidth storage space. Error-bounded lossy compression has been considered one most effective solutions to above problem. In practice, however, best-fit method often needs be customized or optimized in particular diverse characteristics different datasets various user requirements on quality performance. this paper, we address issue with novel...
Blind image quality assessment (BIQA) aims to auto-matically evaluate the perceived of a single image, whose performance has been improved by deep learning-based methods in recent years. However, paucity labeled data somewhat restrains BIQA from unleashing their full potential. In this paper, we propose solve problem pretext task customized for self-supervised learning manner, which enables representations orders mag-nitude more data. To constrain process, quality-aware contrastive loss...
Many scientific applications opt for particles instead of meshes as their basic primitives to model complex systems composed billions discrete entities. Such span a diverse array domains, including molecular dynamics, cosmology, computational fluid and geology. The scale the in those increases substantially thanks ever-increasing power high-performance computing (HPC) platforms. However, actual gains from such are often undercut by obstacles data management related storage, transfer,...
Millions of people in the United States experience a reduced or distorted ability to smell taste. Chemosensory disorders such as anosmia (the inability smell), parosmia (distorted dysgeusia (altered taste) have major impacts on health and quality life including difficulty sensing dangers fire spoilage, diminished palatability food drink that can negatively influence diet nutrition, feelings social isolation, an increased incidence frailty, anxiety, depression. Smell taste dysfunction also be...
In this paper, full bottom dielectric isolation (BDI) is first demonstrated on horizontally stacked Nanosheet device structures with Lmetal 12 nm. The comparison of BDI scheme vs punch through stopper (PTS) has been systematically studied. By comparing off-state leakage current, short channel behavior and effective capacitance (Ceff) for both schemes, we show that could potentially provide: 1) good immunity sub-channel due to process variation (from parasitic "fat-Fin" which unique in...
Efficient error-controlled lossy compressors are becoming critical to the success of today's large-scale scientific applications because ever-increasing volume data produced by applications. In past decade, many lossless and have been developed with distinct design principles for different datasets in largely diverse domains. order support researchers users assessing comparing a fair convenient way, we establish standard compression assessment benchmark -- Scientific Data Reduction Benchmark...
Today's extreme-scale high-performance computing (HPC) applications are producing volumes of data too large to save or transfer because limited storage space and I/O bandwidth. Error-bounded lossy compression has been commonly known as one the best solutions big science issue, it can significantly reduce volume with strictly controlled distortion based on user requirements. In this work, we develop an adaptive parameter optimization algorithm integrated a series strategies for SZ,...
Experimental reliability trends indicate that t <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">inv</sub> -scaling with HKMG stacks remains challenging because NBTI, PBTI and TDDB margins rapidly decrease decreasing values increasing gate leakage current. A case is made these observed arise from the layer structure materials properties of SiO(N)/HfO xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> dual dielectric. Therefore, fundamental...
This paper reports on the NTIRE 2023 Quality Assessment of Video Enhancement Challenge, which will be held in conjunction with New Trends Image Restoration and Workshop (NTIRE) at CVPR 2023. challenge is to address a major field video processing, namely, quality assessment (VQA) for enhanced videos. The uses VQA Dataset Perceptual (VDPVE), has total 1211 videos, including 600 videos color, brightness, contrast enhancements, 310 deblurring, 301 deshaked 167 registered participants. 61...
We present a fully integrated 7nm CMOS platform featuring 3 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">rd</sup> generation finFET architecture, SAQP for fin formation, and SADP BEOL metallization. This technology reflects an improvement of 2.8X routed logic density >40% performance over the 14nm reference described in [1-3]. A full range Vts is enabled on-chip through unique multi-workfunction process. enables both excellent low voltage...
Error-bounded lossy compression is becoming an indispensable technique for the success of today’s scientific projects with vast volumes data produced during simulations or instrument acquisitions. Not only can it significantly reduce size, but also control errors based on user-specified error bounds. Autoencoder (AE) models have been widely used in image compression, few AE-based approaches support error-bounding features, which are highly required by applications. To address this issue, we...
Video quality assessment (VQA) aims to simulate the human perception of video quality, which is influenced by factors ranging from low-level color and texture details high-level semantic content. To effectively model these complicated quality-related factors, in this paper, we decompose into three levels (i.e., patch level, frame clip level), propose a novel Zoom-VQA architecture perceive spatio-temporal features at different levels. It integrates components: attention module, pyramid...
Error-bounded lossy compression has been identified as a promising solution for significantly reducing scientific data volumes upon users' requirements on distortion. For the existing error-bounded compressors, some of them (such SPERR and FAZ) can reach fairly high ratios others SZx, SZ, ZFP) feature speeds, but they rarely exhibit both ratio speed meanwhile. In this paper, we propose HPEZ with newly-designed interpolations quality-metric-driven auto-tuning, which features improved quality...
GPU-aware collective communication has become a major bottleneck for modern computing platforms as GPU power rapidly rises. A traditional approach is to directly integrate lossy compression into collectives, which can lead serious performance issues such underutilized devices and uncontrolled data distortion. In order address these issues, in this paper, we propose gZCCL, first-ever general framework that designs optimizes GPU-aware, compression-enabled collectives with an accuracy-aware...
In this paper, fundamental aspects of the Bias Temperature Instability (BTI) in FETs with metal gate/high-k (HKMG) gate stacks are discussed from a single defect point view. First, Random Telegraph Noise (RTN) measurements used to show that capture/emission processes individual defects highly scaled HKMG exhibit very similar Poisson statistics and can be fully characterized by characteristic electron/hole capture, τ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML"...
Linear algebra operations have been widely used in big data analytics and scientific computations. Many works done on optimizing linear GPUs with regular-shaped input. However, few are focusing fully utilizing GPU resources when the input is not regular-shaped. Current optimizations lack of considering memory bandwidth computing power, therefore they could only achieve sub-optimal performance. In this paper, we propose a performant tall-and-skinny matrix-matrix multiplication algorithm -...
Convolutional neural networks (CNNs) are becoming more and important for solving challenging critical problems in many fields. CNN inference applications have been deployed safety-critical systems, which may suffer from soft errors caused by high-energy particles, high temperature, or abnormal voltage. Of importance is ensuring the stability of process against errors. Traditional fault tolerance methods not suitable because error-correcting code unable to protect computational components,...
Stacked Gate-All-Around (GAA) nanosheet pFETs with compressively strained Si <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1-x</sub> Ge xmlns:xlink="http://www.w3.org/1999/xlink">x</sub> channel have been fabricated to explore their electrical benefits. The NS structure high crystalline quality and 1GPa compressive stress has realized for the first time. Systematic study performed understand effect of epitaxial thickness, fraction, cap...
&#160; As climate and weather scientists strive to increase accuracy understanding of our world, models have increased in their resolution square kilometers scale become more complex increasing demands for data storage. A recent study SCREAM run at 3.5km produced nearly 4.5TB per simulated day, the CMIP6 simulations 28PB data. At same time, storage power capacity facilities conducting experiments are not rate as volume datasets leading a pressing challenge reduce volumes. While some...
Increasing data volumes from scientific simulations and instruments (supercomputers, accelerators, telescopes) often exceed network, storage, analysis capabilities. The community's response to this challenge is reduction. Reduction can take many forms, such as triggering, sampling, filtering, quantization, dimensionality This report focuses on a specific technique: lossy compression. Lossy compression retains all points, leveraging correlations controlled reduced accuracy. Quality...