THTsen-Wei Hsu

Tsen-Wei Hsu

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Tsen-Wei Hsu has a total of 16 co-authors. Here are the top co-authors by number of shared publications:

  • B. N. Lin

    Department of Physics, National Tsing Hua University, Hsinchu 300, Taiwan

    1 shared publication
  • C. H. Chang

    Department of Physics, National Central University, Chungli 320, Taiwan

    1 shared publication
  • C. H. Lin

    Department of Chemistry, Tamkang University, Tamsui 251, Taiwan

    1 shared publication
  • H. C. Chiang

    Department of Physics, National Tsing Hua University, Hsinchu 300, Taiwan

    1 shared publication
  • H. C. Ku

    Department of Physics, National Tsing Hua University, Hsinchu 300, Taiwan

    1 shared publication
  • H. M. Luo

    Department of Physics, National Tsing Hua University, Hsinchu 300, Taiwan

    1 shared publication

Latest Publications

  • Open Access

    SG-Float: Achieving Memory Access and Computing Power Reduction Using Self-Gating Float in CNNs

    by Jun-Shen Wu, Tsen-Wei Hsu, Ren-Shuo Liu

    Convolutional neural networks (CNNs) are essential for advancing the field of artificial intelligence. However, since these networks are highly demanding in terms of memory and computation, implementing CNNs can be challenging. To make CNNs more accessible to energy-constrained devices, researchers are exploring new algorithmic techniques and hardware designs that can reduce memory and computation requirements. In this work, we present self-gating float (SG-Float), algorithm hardware co-design of a novel binary number format, which can significantly reduce memory access and computing power requirements in CNNs. SG-Float is a self-gating format that uses the exponent to self-gate the mantissa to zero, exploiting the characteristic of floating-point that the exponent determines the magnitude of a floating-point value and the error tolerance property of CNNs. SG-Float represents relatively small values using only the exponent, which increases the proportion of ineffective mantissas, corresponding to reducing mantissa multiplications of floating-point numbers. To minimize the accuracy loss caused by the approximation error introduced by SG-Float, we propose a fine-tuning process to determine the exponent thresholds of SG-Float and reclaim the accuracy loss. We also develop a hardware optimization technique, called the SG-Float buffering strategy, to best match SG-Float with CNN accelerators and further reduce memory access. We apply the SG-Float buffering strategy to vector-vector multiplication processing elements (PEs), which NVDLA adopts, in TSMC 40nm technology. Our evaluation results demonstrate that SG-Float can achieve up to 35% reduction in memory access power and up to 54% reduction in computing power compared with AdaptivFloat, a state-of-the-art format, with negligible power and area overhead. Additionally, we show that SG-Float can be combined with neural network pruning methods to further reduce memory access and mantissa multiplications in pruned CNN models. Overall, our work shows that SG-Float is a promising solution to the problem of CNN memory access and computing power.


  • Variation of anomalous Pr ordering and crystal symmetry for the oxygenated Pr1+xBa2−xCu3O7+y system

    by H. M. Luo, B. N. Lin, Y. H. Lin, H. C. Chiang, Y. Y. Hsu, T. I. Hsu, T. J. Lee, H. C. Ku, C. H. Lin, H.-C. I. Kao, et al.

    Structural, transport, magnetic, calorimetric, and neutron data were reported for the oxygen-annealed Pr1+xBa2−xCu3O7+y or 1212-type Cu(Ba2−xPrx)PrCu2O7+y system (0<~x<~1, −0.11<~y<~0.31). Powder x-ray Rietveld analysis indicates that due to subtle oxygen distribution in the CuO1+y plane for these oxygenated cuprates, two structural symmetry transitions were observed, from orthorhombic 1212C (CuO chain) O(I) phase (space group Pmmm) for 0<~x<~0.4, to tetragonal 1212 T phase (P4/mmm) for 0.4<~x<~0.65, and then to a different type of orthorhombic 1212 O(II) phase (Cmmm) for 0.65<x<1. Resistivity data indicate that Pr-in-Ba is not favorable for metallic state and no superconductivity can be detected for these insulating cuprates. Magnetic-susceptibility, heat-capacity, and neutron-diffraction data show that, regardless of the structural transitions, the anomalous Néel temperature TN(Pr) decreases monotonically and smoothly from 18 K for x=0 to 2.5 K for x=0.8, with the same c-axis antiferromagnetic Pr spin alignment. The increasing Pr-O bond length observed with decreasing TN(Pr) indicates that this unusual Pr magnetic ordering is closely correlated with the wave-function overlap between Pr−4f orbital and eight O−2pπ orbital in the CuO2 bilayer.