Cyber Risks · Trail of Bits Blog

Ryan Eberhardt

Summary: Trail of Bits published a blog post by Ryan Eberhardt of Stanford University on October 22, 2020, describing preliminary work on using GPUs to fuzz embedded software in the cloud. The post states that fuzzing supplies programs with many randomized inputs to cause unexpected behavior, and that based on this preliminary work the team believes GPUs can deliver 10x performance per dollar.

Bottom line: The claim is 10x performance per dollar for fuzzing embedded software in the cloud, which is where the cost of finding bugs in that class of software is set.

Read the original at Trail of Bits Blog ↗