Spring '26
ECE 788 - Hardware Acceleration for Machine Learning (3 credits) New Jersey Institute of Technology
Course Objectives: This course provides an introduction to machine learning hardware systems, focusing on the interaction between modern learning algorithms and their underlying computing architectures. Students will study core machine learning models and examine how their computational characteristics influence hardware design. The course explores hardware/software co-design challenges, data mapping strategies, and implementation considerations for efficient machine learning acceleration. Topics include acceleration platforms such as GPUs, systolic arrays, and FPGAs, along with industry-standard benchmarking methodologies such as MLPerf. The course also introduces emerging computing paradigms, including analog crossbar architectures and in-memory computing, highlighting their potential to overcome energy and performance limitations of conventional architectures.
Spring '22 - '26
ECE 452 - High Performance Computer Architecture (3 credits) New Jersey Institute of Technology
Course Objectives: This course explores advanced and emerging topics in computer architecture with a focus on machine learning–centric computing systems and memory-centric challenges. Students study modern neural network classifiers, including multilayer perceptrons (MLPs), convolutional neural networks (CNNs), and spiking neural networks (SNNs), and examine the architectural implications of their computational and dataflow characteristics. The course covers hardware architectures for machine learning acceleration, including accelerator design metrics, hardware/software co-design challenges, data mapping strategies, and computation reduction techniques. In addition, the course investigates memory system security and reliability, with emphasis on DRAM vulnerabilities such as RowHammer and RowPress, DRAM testing infrastructures, and advanced memory technologies. The course also introduces intrinsically pipelined and non-von Neumann computing paradigms, including lateral spin valve–based systems and quantum-dot cellular automata, highlighting their potential for future energy-efficient and scalable architectures.
Fall '21- '25
ECE 451 - Advanced Computer Architecture (3 credits) New Jersey Institute of Technology
Course Objectives: This course focuses on advanced concepts in computer systems design, and the interaction between hardware and software components at various levels (i.e., hardware/software codesign). It introduces common performance measures and tradeoffs used by hardware and software designers to facilitate comparative analysis. The main topics are power wall and memory wall technology challenges, pipelining, multicore architecture, advanced memory technologies with an emphasis on non-volatile memories, introduction to parallel computing, domain-specific architectures (i.e., FPGA, ASIC), and an introduction to analog and digital in-memory computing.
Fall '19
EEL 4362: Post-CMOS Devices and Circuits (3 credits) University of Central Florida
Course Objectives: To introduce state-of-the-art complementary metal-oxide-semiconductor (CMOS) technology and post-CMOS technologies including spintronic, Quantum-dot Cellular Automata, memristor, tunneling FET devices, and their applications in emerging memory, logic, and neuromorphic computing.