Norihisa Fujita

Learn More
G-protein-coupled receptor (GPCR) is one of the most important targets for medicines. Homology modeling based on the crystal structure of bovine rhodopsin is currently the most frequently used method for GPCR targeted drug design. Information about residue-residue contacts and the structural specificity in the subfamily is essential for constructing more(More)
G-protein-coupled receptors (GPCRs) play a crucial role in signal transduction and receive a wide variety of ligands. GPCRs are a major target in drug design, as nearly 50% of all contemporary medicines act on GPCRs. GPCRs are membrane proteins possessing a common structural feature, seven transmembrane helices. In order to design an effective drug to act(More)
Inter-node communications between accelerators in heterogeneous clusters require extra latency because of the time required to transfer data copies between the host and accelerator. Such communication latencies inhibit the optimal performance of affected applications. To address this problem, we proposed the Tightly Coupled Accelerators (TCA) architecture(More)
The Tightly Coupled Accelerators (TCA) architecture that we proposed in previous work enables direct communication between accelerators over nodes. In this paper, we present a proof-of-concept GPU cluster called the HA-PACS/TCA using the PEACH2 chip that we designed as an interconnection router chip based on the TCA architecture. Our system demonstrated 2.0(More)
The conservative global gyrokinetic toroidal full-f five-dimensional Vlasov simulation (GT5D) is a nuclear fusion simulation program designed to analyze turbulence phenomena in tokamak plasma. In this research, we optimize it for graphics processing unit (GPU) clusters with multiple GPUs on each node. Based on the profile results of a GT5D on a CPU node, it(More)
GT5D is a nuclear fusion simulation program which aims to analyze the turbulence phenomena in tokamak plasma. In this research, we optimize it for GPU clusters with multiple GPUs on a node. Based on the profile result of GT5D on a CPU node, we decide to offload the whole of the time development part of the program to GPUs except MPI communication. We(More)
  • 1