, C. D. Peters-Lidard

to perform global land surface simulations at down to 1-km resolution in near-realtime. Such an unprecedented scale and intensity pose many challenges in computational technology. In this report we will demonstrate our developments and innova-tions in high performance computing to meet these challenges and reach our performance goals. These contributions include a fault-tolerant job management system running on a Linux cluster, high-performance, high-availability parallel IO based on GrADS-DODS (GDS) servers with dynamic load-balancing and distributed data storage, and highly scalable data replication with peer-to-peer (P2P) technology. These developments are critical in making LIS a high performance Earth System modeling component with broader applications, and made LIS suitable for production runs in Grid computing environments, and readily interoperable with other modeling components over the Internet. I