The relative cost of memory operations continues to increase as processor performance improvements outpace memory performance improvements. In operating systems based upon copy semantics, this widening performance gap represents an ever-increasing portion of application execution time consumed by I/Ooverhead. Buffering schemes such as Fbufs and Container Shipping incorporate virtual memory page remapping and exploit locality in I/O traffic in an effort to reduce memory copying in the kernel. This thesis introduces a circular (ring) buffer management facility that reduces computation on the I/O critical path and allows greater control over memory consumed in the course of I/O. The system is implemented in Linux and cross-domain and network I/O bandwidths are measured, analyzed, and compared to those offered by Fbufs.