This graph shows the utilization of the Earth Simulator processor nodes (the average value per week), which excludes scheduled maintenance. The whole system is taken out of service once every 2 months for preventive maintenance and once a year for facilities checkups for Yokohama Institute. 16 nodes are out of service once a week for preventive maintenance.
- "running" ... running requests
- "Pre-run/post-run processing" ... "stage-in/stage-out files to work disks, or waiting until necessary nodes are available
- "idle" ... waiting for requests
- "stopped" ... out of operation other than scheduled maintenance