Association for Computing Machinery
In this paper, the authors explore technologies to achieve low-latency energy efficient communications in global scale cloud computing systems. A global-scale cloud computing system linking 100 remote data-centers can interconnect potentially 5M servers, considerably larger than the size of traditional High-Performance Computing (HPC) machines. Traditional HPC machines use tightly coupled processors and networks which rarely drop packets. In contrast, today's IP Internet is a relatively loosely-coupled best-effort network with poor latency and energy-efficiency guarantees, with relatively high packet loss rates.