There is a move toward multi-core processor architectures in server designs. Application software can exploit this capability to attain higher performance by using thread-level parallelism. System software, however, cannot attain a significant performance gain using this strategy. For example, server network processing performance suffers due to poorly scaling high latency operations such as I/O bus crossings, cache misses and interrupts. Multi-core architectures enable a different strategy to boost server system software performance.

Teja NP is a framework for enabling system software scalability and performance improvement on multi-core processors. The key focus of this framework is the use of parallel and pipeline execution techniques for retargeting system software on multi-core architectures. Benchmark results using Teja NP show significant I/O performance improvement in servers. Teja NP is a well-accepted design methodology in the communications market and it has the potential to create disruptive performance gain in the data center and enterprise computing market.