Accelerating Microarchitectural Security Research

Microarchitectural security can be an intimidating field to get started in. When the effectiveness of a microarchitectural attack or defense rests on a single opcode – which is necessary due to undocumented microarchitectural details that must be painstakingly reverse-engineered – it’s easy to get a false result, or to be discouraged away from the field entirely. Yet the severity of the recent transient execution vulnerabilities (such as Meltdown, Spectre, Foreshadow, MDS, and LVI) has demonstrated how crucial awareness of microarchitectural security is at all layers of the stack in both hardware and software engineering. We are working on processes and tooling to lower the barrier to entry into the field and accelerate attack research and mitigation across academia and industry.