Simulation - Driven Design of High - Performance Programmable Network Interface Cards by Paul Willmann A

Simulation-Driven Design of High-Performance Programmable Network Interface Cards by Paul Willmann As network link speeds race to 10 Gigabit/sec and beyond, Internet servers will rely on programmable network interface cards (NICs) to relieve the ever increasing frame processing burdens. To meet that need, this work introduces a scalable, programmable NIC architecture that saturates a full-duplex 10 Gigabit/sec Ethernet link. This proposed architecture utilizes simple parallel processors instead of a single complex core to satisfy its frame-processing requirements, thereby reducing core power by 63%. To exploit lower-frequency parallel resources, this work also contributes an enhanced event queue firmware mechanism that enables frame-level parallelism. Although simulation provides a detailed, inexpensive method to evaluate architectures and software, no detailed architectural simulator has previously targeted NIC designs. This work therefore contributes Spinach, a new simulation toolset that accurately models programmable NICs in microarchitectural detail. A Spinach model of an existing Gigabit NIC validates hardware benchmarks within 8.9% and yields solutions to previously undiscovered performance bottlenecks. Acknowledgments I would like to thank Dr. Vijay Pai and Dr. Scott Rixner for their motivation and guidance in this work. Also, I would like to thank Dr. Bart Sinclair for his perspectives on simulation approaches and methodology. Additionally, I want to acknowledge Mike Brogioli’s significant technical and moral support throughout the development of this project. Many coffee-sustained nights (and following days) went into this work, and Mike worked through many of them with me. Hyong-youb Kim’s expertise on all things related to the Tigon and its firmware proved invaluable toward my understanding in the development of this work. Neil Vachharajani provided extensive debugging of and improvements to the Liberty backend, for which I am very thankful. I also thank Dr. Jan Hewitt for sharing her vast wisdom regarding language and thesis semantics with me. Finally, I must thank my family, Leighann, and my friends. This work would not be possible without their steadfast support and understanding.