Network I/O Acceleration in Heterogeneous Multicore Processors
HOTI '06 Proceedings of the 14th IEEE Symposium on High-Performance Interconnects
Understanding Linux Network Internals
Understanding Linux Network Internals
TCP offload is a dumb idea whose time has come
HOTOS'03 Proceedings of the 9th conference on Hot Topics in Operating Systems - Volume 9
Studying network protocol offload with emulation: approach and preliminary results
HOTI '04 Proceedings of the High Performance Interconnects, 2004. on Proceedings. 12th Annual IEEE Symposium
A New Offloaded/Onloaded Network Interface for High Performance Communication
PDP '09 Proceedings of the 2009 17th Euromicro International Conference on Parallel, Distributed and Network-based Processing
| Hi-index | 0.00 |
In the last years, the market is demanding (scientific, multimedia, real-time, etc.) applications with high bandwidth requirements. To support this, the bandwidth of the network links has increased to reach multiple gigabit per second. Nevertheless, taking advantage of multigigabit per second links requires a lot of processor cycles for communication tasks, diminishing the processor cyles that remains available for the application. Actual multiprocessor and multicore architectures as well as programmable NICs (Network Interface Cards) provide new opportunities to exploit the parallelism, distributing the communication overhead across the processors available in the node. We have designed a network interface that takes advantage of the different processors available in the node. In this paper, the advantages of this optimized network interface are shown by analyzing the performance of a web server application.