Introductory description

ES4F3-15 High Performance Embedded Systems Design.

Module web page

Module aims

Outline syllabus

This is an indicative module outline only to give an indication of the sort of topics that may be covered. Actual sessions held may differ.

Design Flow and Verilog Recap: Revision of Register Transfer Level (RTL) design, behavioural modelling, hardware synthesis and implementation flow.

Design Space Exploration: Understanding area, performance, and power in the context of embedded systems, and the trade-offs between them.

High Level Synthesis: C-to-gates and other languages, compiler outline, supported language features, basic scheduling.

Interconnect and Data Movement: Processor to logic interfacing, busses and crossbars, networks-on-chip, Direct Memory Access (DMA).

Modern Hybrid FPGA Architecture: Advanced DSP block architecture, advanced I/O, partial reconfiguration, hybrid FPGA architecture, PCIe-connected FPGAs.

Accelerating Algorithms: Standard DSP structures, including FIR filters, image filters, number representation and numerical precision.

Learning outcomes

By the end of the module, students should be able to:

• Apply advanced features of FPGA architectures in high performance embedded systems design.
• Design a hardware accelerator for a complex algorithm by evaluating its parallelism and arithmetic requirements.
• Understand how to integrate a hardware accelerator with a processor and design the necessary software and hardware communication infrastructure.
• Apply specialist practical knowledge of hardware design at the register transfer level and understand the principles of high level synthesis.
• Understand the trade-offs between performance, area, and power in critiquing the design of accelerators.

Indicative reading list

Embedded Systems Design with FPGAs, P Athanas, D. Pnevmatikatos, N. Sklavos (Editors), Springer, 2013.

Embedded Systems Fundamentals with Arm Cortex-M based Microcontrollers, A. G. Dean, Arm Education Media UK, 2017.

The Zynq Book: Embedded Processing with the ARM® Cortex-A9 on the Xilinx Zynq-7000 All Programmable SoC, Louise H. Crockett, Ross A. Elliot, Martin A. Enderwitz, Robert W. Stewart, Strathclyde Academic Media, 2014

Parallel Programming for FPGAs, R. Kastner, J. Matai, and S. Neuendorffer, arXiv, 2018.

Subject specific skills

Ability to conceive, make and realise a component, product, system or process
Ability to be pragmatic, taking a systematic approach and the logical and practical steps necessary for, often complex, concepts to become reality

Transferable skills

Communicate (written and oral; to technical and non-technical audiences) and work with others
Exercise initiative and personal responsibility, including time management, which may be as a team member or leader
Overcome difficulties by employing skills, knowledge and understanding in a flexible manner