Ph.D.

Research subject

Design of probabilistic real-time systems

Director/codirector

Director: Giovanni Beltrame  

Posting

End of posting : 20 June 2013

Units and departments

Department of Computer Engineering, MIST Lab

Conditions

GPA >= 3.5, some knowledge of probability and statistics, experience in hardware design (VHDL and Verilog) is a plus

Application

To apply: http://mistlab.ca/join

Detailed description

Background

One of the main obstacles to high-performance aerospace computer systems is the unpredictable timing behaviour of modern computer architectures; with multi-stage pipelines and three-level memory hierarchies, it is extremely difficult to accurately predict the execution time of a given program. This is made almost impossible with parallel (such as multi-core) architectures, due to the presence of shared resources. Being conservative and over-estimating does not solve the problem because of the extremely large difference between worst and average cases. The design of powerful real-time systems needs a new approach for the architecture of instruction processors. In this context, a probabilistic analysis approach is proposed : by enabling true randomized behaviour in all the components of computer, one
can define probabilistic metrics to the timing behaviour of a system. Successful implementation of such systems will have tremendous impact on the way critical systems are designed, and the potential benefits in terms of cost of integration, verification, and certification of real-time software are enormous.

Objectives

• Evaluation of current probabilistic approaches
• Implementation of a probabilistic computer architecture
• Validation on realistic applications

Possible funding

Financial support is possible

NSERC subjects

• 2700 INFORMATION TECHNOLOGY
• 2702 Computer systems organization
• 2719 Computer architecture and design