K
Koichi Shirahata
Researcher at Fujitsu
Publications - 18
Citations - 192
Koichi Shirahata is an academic researcher from Fujitsu. The author has contributed to research in topics: General-purpose computing on graphics processing units & Multi-core processor. The author has an hindex of 7, co-authored 16 publications receiving 172 citations. Previous affiliations of Koichi Shirahata include Tokyo Institute of Technology.
Papers
More filters
Proceedings ArticleDOI
Hybrid Map Task Scheduling for GPU-Based Heterogeneous Clusters
TL;DR: A hybrid scheduling technique for GPU-based computer clusters, which minimizes the execution time of a submitted job using dynamic profiles of Map tasks running on CPU cores and GPU devices is proposed.
Proceedings ArticleDOI
Out-of-core GPU memory management for MapReduce-based large-scale graph processing
TL;DR: This work proposes a MapReduce-based out-of-core GPU memory management technique for processing large-scale graph applications on heterogeneous GPU-based supercomputers and studies the performance characteristics of the technique, including application's performance and power efficiency of scale-up and scale-out approaches.
Proceedings ArticleDOI
Memory reduction method for deep neural network training
TL;DR: A method to reduce the amount of memory for training a deep neural network is presented, which enables to suppress memory increase during the backward pass, by reusing the memory regions allocated for the forward pass.
Proceedings ArticleDOI
A scalable implementation of a mapreduce-based graph processing algorithm for large-scale heterogeneous supercomputers
TL;DR: This work implemented a multi-GPU-based GIM-V application with load balance optimization between GPU devices that extends the existing MapReduce library for supporting multi- GPU-environments using the MPI library and optimizes load balance betweenGPU devices by employing task scheduling-based graph partitioning.
Proceedings ArticleDOI
Speed-Up of Object Detection Neural Network with GPU
Takuya Fukagai,Kyosuke Maeda,Satoshi Tanabe,Koichi Shirahata,Yasumoto Tomita,Atsushi Ike,Akira Nakagawa +6 more
TL;DR: A speed-up of an object detection neural network with GPU that can detect objects with 55.8ms per image and is also applicable to other detection networks such as R-FCN, YOLO and SSD.