• Chinese Optics Letters
  • Vol. 18, Issue 7, 070901 (2020)
Hiromi Sannomiya1, Naoki Takada2、*, Kohei Suzuki1, Tomoya Sakaguchi1, Hirotaka Nakayama3, Minoru Oikawa2, Yuichiro Mori2, Takashi Kakue4, Tomoyoshi Shimobaba4, and Tomoyoshi Ito4
Author Affiliations
  • 1Graduate School of Integrated Arts and Sciences, Kochi University, Kochi 780-8520, Japan
  • 2Research and Education Faculty, Kochi University, Kochi 780-8520, Japan
  • 3National Astronomical Observatory of Japan, Mitaka 181-8588, Japan
  • 4Graduate School of Engineering, Chiba University, Inage-ku 263-8522, Japan
  • show less
    DOI: 10.3788/COL202018.070901 Cite this Article Set citation alerts
    Hiromi Sannomiya, Naoki Takada, Kohei Suzuki, Tomoya Sakaguchi, Hirotaka Nakayama, Minoru Oikawa, Yuichiro Mori, Takashi Kakue, Tomoyoshi Shimobaba, Tomoyoshi Ito. Real-time spatiotemporal division multiplexing electroholography for 1,200,000 object points using multiple-graphics processing unit cluster[J]. Chinese Optics Letters, 2020, 18(7): 070901 Copy Citation Text show less
    Spatiotemporal division multiplexing approach for suppressing the deterioration of a 3D holographic video reconstructed from a point-cloud model comprising a huge number of object points.
    Fig. 1. Spatiotemporal division multiplexing approach for suppressing the deterioration of a 3D holographic video reconstructed from a point-cloud model comprising a huge number of object points.
    Spatiotemporal division multiplexing approach using moving image features.
    Fig. 2. Spatiotemporal division multiplexing approach using moving image features.
    Reconstructed 3D image from a 3D object “fountain” comprising 1,064,462 object points.
    Fig. 3. Reconstructed 3D image from a 3D object “fountain” comprising 1,064,462 object points.
    Multi-GPU cluster system with multiple GPUs connected by a gigabit Ethernet network and a single SLM.
    Fig. 4. Multi-GPU cluster system with multiple GPUs connected by a gigabit Ethernet network and a single SLM.
    Pipeline processing for the spatiotemporal electroholography system shown in Fig. 2.
    Fig. 5. Pipeline processing for the spatiotemporal electroholography system shown in Fig. 2.
    Read data processing and CGH calculation on each CGH calculation node in the multi-GPU cluster system shown in Fig. 4. (a) Serial computing. (b) Parallel computing.
    Fig. 6. Read data processing and CGH calculation on each CGH calculation node in the multi-GPU cluster system shown in Fig. 4. (a) Serial computing. (b) Parallel computing.
    Comparison of the total display time for every 12 frames using serial computing shown in Fig. 6(a) with that using parallel computing shown in Fig. 6(b) when the number of object points is 1,200,000.
    Fig. 7. Comparison of the total display time for every 12 frames using serial computing shown in Fig. 6(a) with that using parallel computing shown in Fig. 6(b) when the number of object points is 1,200,000.
    Display-time interval T shown in Fig. 5 plotted versus the number of object points when using the spatiotemporal division multiplexing approach using moving image features implemented on the multi-GPU cluster system shown in Fig. 4.
    Fig. 8. Display-time interval T shown in Fig. 5 plotted versus the number of object points when using the spatiotemporal division multiplexing approach using moving image features implemented on the multi-GPU cluster system shown in Fig. 4.
    Snapshot of a reconstructed 3D video (Video 1).
    Fig. 9. Snapshot of a reconstructed 3D video (Video 1).
    CPUIntel Core i7 7800X (clock speed: 3.5 GHz)
    Main memoryDDR4-2666 16 GB
    OSLinux (CentOS 7.6 x86_64)
    SoftwareNVIDIA CUDA 10.1 SDK, OpenGL, MPICH 3.2
    GPUNVIDIA GeForce GTX 1080 Ti
    Table 1. Specifications of Each Node in the Multi-GPU Cluster System
    Number of Space DivisionsObject PointsFrame Rate (fps)
    No division1,064,4625.43
    Two divisions532,23110.86
    Four divisions266,11621.70
    Six divisions177,41132.70
    Table 2. Frame Rate of the Reconstructed 3D Video from the Original 3D Video “Fountain” Comprising 1,064,462 Object Points Against the Number of Space Divisions
    Hiromi Sannomiya, Naoki Takada, Kohei Suzuki, Tomoya Sakaguchi, Hirotaka Nakayama, Minoru Oikawa, Yuichiro Mori, Takashi Kakue, Tomoyoshi Shimobaba, Tomoyoshi Ito. Real-time spatiotemporal division multiplexing electroholography for 1,200,000 object points using multiple-graphics processing unit cluster[J]. Chinese Optics Letters, 2020, 18(7): 070901
    Download Citation