2022 Fiscal Year Research-status Report
Large-scale Tomography Computation
| Project/Area Number |
21K17750
|
|---|
| Research Institution | National Institute of Advanced Industrial Science and Technology |
|---|
Principal Investigator |
陳 鵬 国立研究開発法人産業技術総合研究所, 情報・人間工学領域, 研究員 (30890199)
|
|---|
| Project Period (FY) |
2021-04-01 – 2025-03-31
|
| Keywords | Computed Tomography / Image Reconstruction / Filtered Back-projection / CPU / GPU / FPGA / Ptychography |
|---|
| Outline of Annual Research Achievements |
We present a novel approach to address the memory and computation challenges in ptychographic image reconstruction, a high-resolution microscopic imaging technique. Our method involves decomposing image gradients and diffraction measurements into tiles, resulting in a reduced memory footprint. Furthermore, we propose a parallel processing technique that facilitates efficient communication and parallel pipelining across a large number of GPUs, such as up to 4000 GPUs.
In addition to GPU acceleration, we explore the use of FPGA accelerators to achieve improved computing and energy efficiency in Computed Tomography image reconstruction.
|
| Current Status of Research Progress |
Current Status of Research Progress
1: Research has progressed more than it was originally planned.
Reason
We propose a parallel processing technique that enables efficient communication and parallel pipelining across a significant number of GPUs, including up to 4000 GPUs. The success and findings of our approach have been published in top-tier conferences.
|
| Strategy for Future Research Activity |
We aim to optimize the computing kernels by leveraging heterogeneous computing architectures, such as the FPGA accelerator, to achieve both computing efficiency and power efficiency
|
| Causes of Carryover |
Previously, student stipends were provided for organizing image reconstruction data. However, this fiscal year had fewer experiments requiring large-scale computations, prompting the research project leader to handle the task independently. Consequently, no stipends were given for assisting with image reconstruction and documentation organization. Moreover, due to COVID, international conferences were conducted online, resulting in no associated expenses. In the upcoming fiscal year, there are plans to attend international conferences in person and conduct more experiments with research assistants.
|
