November 16th, 2005
Geoscientists and bioscientists have a need to seamlessly acquire, transfer, visualize and analyze terabytes of data using advanced multi-gigabit networks.
For the StorCloud experiment, EVL researchers and collaborators propose to transfer large high-resolution and time-varying 2D and 3D geophysical and biological datasets, on demand, from remote storage servers located in San Diego, Chicago and Amsterdam to the StorCloud storage in Seattle. National and international advanced research optical networks such as CAVEwave, TransLight, SURFnet, CA*net4 and the National LambdaRail will be used.
The datasets will be visualized onsite using Ethereon, an interactive environment for visual exploration of multi-resolution datasets.
The visualization will be showcased on a high-resolution tiled-display system driven by a cluster on the SC05 show floor.
Ethereon walks through terabyte datasets using LambdaRAM, a data middleware that fetches and prefetches data from multiple heterogeneous storage systems. We will use LambdaStream, a user space, multi-flow multi-cluster data transport protocol for Long Fat Networks, to transfer data from remote data servers to the StorCloud storage.