ECE 69500 - Programmable Accelerator ArchitecturesLecture Hours: 3 Credits: 3
Areas of Specialization(s):
Experimental Course Offered: Spring 2018, Spring 2020
Requisites by Topic:
This class builds on previous knowledge of general-purpose processor design to explore the space of programmable hardware accelerators. Hardware accelerators seek to fulfill the promise of continued performance and energy-efficiency gains in the era of a slowing Moore's law, larger problem sizes and an increased focused on energy-efficiency. These factors have caused hardware acceleration to become ubiquitous in today's computing world and critically important in computing's future. This class will introduce students to the architectures of programmable accelerators. We will delve deeply into the architectures of modern massively parallel accelerators like GPUs, culminating in a course project using an open-source research and development simulator used in academia and industry. General topics in hardware acceleration will be discussed, including but not limited to GPGPU and massively parallel computing, approximate accelerators, reconfigurable hardware and programmable hardware for machine learning.
- General Purpose Graphics Processor Architecture, Aamodt, T.M., Fung, W.L., & Rogers, T.G,, 2018.
- Computer Architecture: A Quantitative Approach, 5th Edition, Hennessey and Patterson.
- Programming Massively Parallel Processors: A Hands-on Approach, 3rd Edition, Kirk, D.B., & Hwu, W.M.W., Elsevier, Inc., 2016.
|1 week||General purpose architecture background and the evolution to accelerators Entropy and channel capacity from the detection perspective|
|2 weeks||Programming massively parallel accelerators|
|1 week||Advances in the GPU programming model|
|3 weeks||GPU core design|
|2 weeks||GPU memory system and interconnect|
|2 weeks||CPU/GPU systems and AMD Fusion architecture|
|1 week||Intel Xeon Phi design|
|1 week||Custom and reconfigurable accelerators|
|2 weeks||Case studies on architectures for machine learning|