Large-scale electroholography by HORN-8 from a point-cloud model with 400,000 points

Area of Science:

• Computer graphics
• Holography
• High-performance computing

Background:

• Computer-generated holograms (CGH) are crucial for advanced display technologies.
• Existing methods often face limitations in processing speed and memory requirements for complex holographic scenes.

Purpose of the Study:

• To develop a high-speed system for generating computer-generated holograms.
• To overcome the computational and memory bottlenecks in current electroholography.

Main Methods:

• Development of the HORN-8 system, a cluster utilizing eight HORN-8 boards.
• Implementation of spatiotemporal division for processing large point-cloud models (~400,000 points).
• Evaluation of holographic video reproduction capabilities.

Main Results:

• The HORN-8 system achieves speeds 1,000 times faster than a personal computer.
• Real-time holographic updates (60 fps) for 2-million-pixel displays from 65,536-point models.
• Successful reproduction of video-holography using ~400,000 points via spatiotemporal division.

Conclusions:

• The HORN-8 system offers a significant advancement in high-speed electroholography.
• The specialized hardware efficiently handles large datasets without requiring extensive internal memory.
• This technology enables high-fidelity holographic video generation.