The Fuzz Odyssey: A Survey on Hardware Fuzzing Frameworks for Hardware Design Verification

Abstract

Hardware Security is at stake driven by the growing complexity and integration of processors, SoCs, and diverse third-party intellectual property (IP) hardware, all geared toward delivering advanced solutions. To preserve the system integrity and mitigate the post-production re-engineering costs, the Design Verification (DV) community employs dynamic and formal verification strategies. However, with the ever-increasing complexity of modern processors, these techniques fall in short of scalability and increased verification time. Recently, hardware fuzzing inspired by software testing has been navigating uncharted territories in hardware bug detection capabilities. Multiple hardware fuzzing techniques have been recently introduced that either utilize the hardware design in its inherent form for fuzzing or convert the hardware into software models and perform fuzzing to detect bugs. However, the existing techniques claim to be a silver bullet in their way, we provide some critical insights on these techniques by reviewing the fundamental principles of hardware fuzzing frameworks, the methodologies involved, and the diverse hardware designs in which they can be employed. Furthermore, we discuss the challenges and limitations of the fuzzing framework. We also present feasible future research directions based on our observations and insights.

Topics & Concepts