This talk will present two microarchitectures that overcome the drawbacks of a traditional LSQ by moving pieces of its functionality out of the out-of-order core and into the in-order front- and back-end pipelines. The first design, SQIP, relies on accurate memory dependence prediction and re-execution based load verification to eliminate associative search from the LSQ. The second design, NoSQ, exploits speculative memory bypassing (SMB) to remove the LSQ entirely.
This is joint work with Tingting Sha and Prof. Milo Martin.
Bio:
Amir Roth is an assistant professor at the Computer and Information Sciences Department at the University of Pennsylvania, where he co-leads the architecture and compilers group (ACG). His current research focuses on scalable microarchitectures and memory systems for multi-core processors. His current projects include minigraph processors (with Anne Bracy), ON-Core (with Andrew Hilton) and NoSQ (with Tingting Sha and Prof. Milo Martin).
Prof. Roth received a BS in physics from Yale University in 1994 and a PhD in computer science from the University of Wisconsin-Madison in 2001. He received the NSF CAREER award in 2002.