Topic domain: Empirical system reliability.

We'll be soon running computers in an environment which is way outside their normal operational parameters. Are the parameters too narrow? What does it really mean that a computer is working correctly?

Evaluate existing correctness benchmarks and assemble a test pattern which tries out as many subsystems as possible. Make sure that not only caches are tested, but entire storage subsystems etc.

Skills:

• Practical, system builder approach
• Computer Organization and Architecture
• Previous experience definitive plus

Focus gauge:

• Theory 2/5: Must understand how computers are actually built, not just in theory
• Coding / Implementation 3/5: most probably just scripting, but the toolkits may be in different languages
• Reading 2/5: Easily readable hardware review sites, e.g., Tom's Hardware, AnandTech. Bianca Schroeder's research.