STANFORD UNIVERSITY - SCHOOL OF ENGINEERING
PERFORMANCE ENGINEERING OF COMPUTER SYSTEMS & NETWORKS
EE-384S - Spring 2024
ADMINISTRATIVE INFORMATION
Class Time: Mon/Wed 4:30-5:50pm, Room: Turing Auditorium in Polya Building.
Prof: Nick Bambos (bambos - at - stanford.edu). Office hours: Mon/Wed 6-6:30pm in Room 238, Packard EE Building.
TA: Emi Zeger, (emizeger - at - stanford.edu) EE PhD student. Office hours: Tue 4-5pm, Room 203, Packard EE Building.
COURSE DESCRIPTION
This is a MS/PhD level course on high-performance engineering of computer systems/networks. The course introduces important core methodologies for modeling and optimization of modern computer systems/networks, and applies them to key design problems related to high-performance engineering of such systems.
The introduced network performance modeling, analysis, evaluation methodologies include: stochastic modeling and Markov chains, queueing systems and networks, stochastic simulation, stochastic scheduling and resource allocation, dynamic programming and network control, network optimization, etc.
Sample applications of the above methodologies to computer system/network engineering design include: traffic modeling and congestion management, admission control and quality of service support, packet scheduling and switching, power control and channel allocation in wireless networks, power management in data centers and sustainable computing, high-reliability robust designs, etc.
We study selected key computer system/network design issues from the previous list, applying the aforementioned methodologies, stressing engineering intuition and developing modeling skills.
Reading material includes lecture notes and slides, book chapters, and research papers from conference proceedings and archival journals.
The homework and class project aim to develop the computer system/network modeling intuition of the student and apply it on important technology design problems in a systematic way.