• 学术搜索
  • 科研智能体
    • Research Labs
    • AI 阅读
    • AI 文库
    • 深度研究
    • 学者亮点
  • 学术资源
    • AI2000
    • 期刊/会议
    • 学者库
    • 学术API
      MCP&Skill
    • 溯源树
    • 数据集
  • 知识沉淀
    • 学术空间
订阅小程序
旧版功能
aminer vip
开通会员低至0.73元/天
一次搞定AI科研
立即登录
  • English
  • 联系方式
    IEEE Open Journal of the Solid-State Circuits Society

    IEEE Open Journal of the Solid-State Circuits Society

    JournalISSN N/AeISSN 2644-1349

    年发文量

    研究主题

    论文(216)

    排序
    1IEEE OPEN JOURNAL OF SOLID-STATE CIRCUITS SOCIETY Special Section on Circuits and Systems for Hardware Security
    Makoto Nagata,Patrick Schaumont,Shivam Bhasin,Yoonmyung Lee
    2026
    引用
    AI阅读
    加入学术空间
    2Guest Editorial Special Section on Advanced Circuits, Architectures, and Systems for Next-Generation AI Processors
    Kea-Tiong Tang,Francesco Conti, Bongjin Kim
    2026
    引用
    AI阅读
    加入学术空间
    3SPEC: A Scalable and Physical-Design-Friendly Architecture for CGRAs
    Po-Han Chen,Alex Carsello,Mark Horowitz,Priyanka Raina

    Modern computing systems must handle increasingly diverse and demanding workloads. Coarse-grained reconfigurable arrays (CGRAs) provide an effective solution by offering flexible datapaths that can be reconfigured at runtime for new applications, while achieving higher performance than CPUs and FPGAs through higher-level operations and spatial parallelism. Recent research has explored many CGRA architectures focusing on flexibility, power efficiency, and performance. However, most of these efforts remain small-scale and limited to RTL exploration, with few advancing to logic synthesis and even fewer achieving silicon implementations. As a result, the scalability, physical design (PD) feasibility, and post-silicon performance of such architectures remain unclear. This paper presents SPEC, a CGRA architecture and hierarchical physical design workflow that enables scalable, physically feasible systems with reliable post-silicon performance. Architecturally, SPEC uses a column-structured tile organization and places global signals and supporting cells inside the macro boundary, forming an abutted, channel-less design that enables compact and scalable physical layouts. This allows the SPEC to use a strictly hierarchical workflow, decomposing the physical design into tile-, array-, and SoC-level stages. With hardened tiles organized hierarchically, even large arrays containing up to 4,096 tiles can complete the physical design flow with clean inter-tile timing in approximately one day on a 72-core server, and yield better timing and smaller areas. In contrast, the conventional thin-channel design spends substantial time optimizing inter-tile timing and eventually times out. To evaluate post-silicon application frequencies, we validate SPEC using three CGRA prototypes implemented with SPEC and report their measured operating frequencies. To quantify the impact of our compiler-aware timing constraints, we compare a representative prototype with a simulated variant in which the constraints are removed. The resulting 9% to 25% degradation in STA frequency demonstrates that SPEC not only supports scalable physical implementation but also enables compiler-mapped applications to operate at high frequencies.

    2026
    引用
    AI阅读
    加入学术空间
    4Transceivers in Recent Commercial Wireless Communication Devices
    Jing-Hong Conan Zhan

    Wireless connectivity is a prerequisite for the artificial-intelligence services that now drive the semiconductor industry, and the radio-frequency (RF) transceivers that realize this connectivity must satisfy standard-driven performance targets under commercial constraints on cost, die area, and process availability. This paper reviews transceiver design considerations for commercial wireless products. The fundamental relations that link signal quality, channel bandwidth, spatial multiplexing, and link budget are first revisited, and the dominant circuit non-idealities are introduced. Commercial considerations are then explained: the importance of long-lasting process nodes, the diminishing power, performance, area, and cost returns from most advanced nodes, and RF properties of active and passive components are discussed, followed by impact due to choice of transceiver partition and interface. Three recently reported transceivers designed for commercial products—a 12-nm CMOS 4G/5G frequency-range-1 (FR1) transceiver with multiple-input multiple-output (MIMO) support, a 28/39-GHz dual-polarized 5G FR2 phased-array transceiver paired with its IF transceiver, and a 55-nm CMOS tri-band switchable Wi-Fi 7 transceiver—are then reviewed as case studies that illustrate how these considerations are resolved in practice.

    2026
    引用
    AI阅读
    加入学术空间
    5Oscillator Phase Noise Theory: A Generalized Impulse Sensitivity Function Framework
    Vahnood Pourahmad,Ehsan Afshari

    The impulse sensitivity function (ISF) framework for the analysis of phase noise is revisited in light of rigorous phase-reduced models (PRMs) of oscillators originally developed in the applied mathematics literature. Starting from a general class of nonlinear oscillator models, we present a systematic methodology for deriving phase-reduced representations and computing the associated ISFs with accuracy and computational efficiency. In particular, the proposed formulation establishes a direct connection between ISFs and the phase sensitivity function (PSF) of the oscillator, which is a more fundamental quantity within the phase-reduction framework, and shows that ISFs can be expressed as certain linear combinations of its components. The proposed approach is applied to three canonical examples, including the Van der Pol–Duffing oscillator and standard Class-B and Class-D differential LC oscillators. Across all phase-noise comparisons presented, the maximum absolute deviation between the proposed approach and SpectreRF is 0.4 dB. Notably, the upconversion of flicker noise to phase noise in standard differential LC CMOS oscillator topologies is explained through the incremental Groszkowski effect and AM–PM conversion, both naturally captured through the oscillator’s phase geometry, without the need for additional phenomenological constructs. This provides a unified, physically grounded, and mathematically rigorous framework for the analysis and efficient computation of phase noise in practical oscillator designs.

    2026
    引用
    AI阅读
    加入学术空间
    立即登录,查看全部 216 篇论文

    高被引作者

    作者引用发文
    Nan Sun1016
    XiYuan Tang993
    Letian Wang811
    Zhijie Ma811
    Jia Jieyang811
    Kai Zang811
    Chao Zhang811
    Jiaxin Liu792
    Linxiao Shen792
    Michael Flynn723

    高产作者

    作者引用发文
    Robert Bogdan Staszewski157
    Nan Sun1016
    Steven M. Bowers466
    Shreyas Sen05
    Teerachot Siriburanon65
    Patrick Mercier555
    Youngcheol Chae225
    Hoi-Jun Yoo474
    Liang Qi104
    Jerald Yoo04

    相关期刊

    暂无数据
    暂无数据