Cross-Layer Reliability of Computing Systems

Computing-system reliability is not a problem confined to one component. Faults can originate in the underlying technology, surface in logic or processor architecture, and eventually affect software and applications. Cross-Layer Reliability of Computing Systems brings those perspectives together in an edited technical volume on designing for resilience and evaluating how well reliability measures work.

Its two-part structure follows a useful progression: first, approaches to improving resilience across system layers; then, methods for assessing reliability, modeling errors, and studying fault behavior.

Reliability from technology to software

The opening part examines the technological, logic, architectural, and software layers. Topics include sources of unreliability in computing technologies, performance and reliability monitors, protection for caches and processor components, software-implemented hardware fault tolerance, and software-based self-test. The organization helps readers see how reliability concerns change as they move from circuits and hardware toward software-level detection and tolerance.

When resilience needs more than one layer

Dedicated coverage of cross-layer resilience considers how techniques at different levels may be combined. This perspective is useful for understanding why a system’s dependability can depend not just on an individual safeguard, but also on how measures across the stack interact.

Assessing faults and their effects

The second part addresses reliability assessment through subjects including physical stress, soft-error modeling and simulation, and fault injection. Its contents also extend to analytical modeling and stochastic methods. Together, these topics give the book a broad technical scope, from experiments and error propagation to approaches for analyzing system behavior.

For readers working with dependable systems

Researchers, postgraduate students, and computer architects focused on computing-system dependability will find a multi-layer reference that connects resilience design with reliability assessment. It is particularly relevant to readers who need to consider faults across technology, circuits, architecture, and software rather than treating each level in isolation.