Researchers in the tri-state region rely on the high-speed connection to MAGPI and Internet2 to enable distributed virtual collaboratory environments and immediate access to very large datasets. At universities around the world researchers collaborate in real time using robust live collaboration technologies. Scientists across disciplines access and distribute very large collections of data for analysis and shared knowledge discovery. New network tools and technologies are being developed by computer scientists, researchers and the R&D community around the globe advancing cyberinfrastructure toward the next generation research and education network and tools for tomorrow.
Discover the possibilities for the Research Community through MAGPI.
Engineers in the Lehigh Valley use live collaboration technologies to perform distributed earthquake simulations. Academics produce and submit content to the Research Channel highlighting research findings across disciplines making access to these discoveries available to thousands of viewers around the world. Physicists take data from the Large Hadron Collider in Geneva and analyze and synthesize that data with their own findings for discoveries into the very origin of the universe. Archeologists collaborate with colleagues in Europe to build a Digital Corinth database of the artifacts of antiquity.
MAGPI and The PHEOBUS Project
Increasing end to end performance for long-distance large data transfers
Conceived and developed by University of Delaware researchers, Phoebus establishes a new network framework and protocol that brings the high performance of advanced backbone networks all the way to the end-user’s desktop. Designed to improve end-to-end through¬put for long-distance data transfers, Phoebus embeds greater “intelligence” in the network, enabling it to choose the best transport technology for any application based on its needs and the available network resources at application run-time. Phoebus works by transparently splitting the full network path into distinct segments at specific adaptation points called “Phoebus gateways” and then finding the best path for the data on a segment-by-segment basis, which could include a combination of IP and dedicated optical circuits. With little to no modification needed by the end-user, applications that utilize Phoebus have expe¬rienced significantly improved throughput.
