We recently proposed a novel mechanism for hardware-based optimization that iteratively optimizes dynamic atomic instruction traces. With iterative optimization, each atomic trace, or frame, is gradually optimized over successive executions. The use of an iterative optimization framework provides the power of dynamic optimization previously demonstrated by frameworks such as rePLay and PARROT via a realizable, table-based, streaming optimizer. In a recent paper submission, we presented the concept of iterative optimization, provided details on the implementation of an iterative optimizer, and provided some data from our experimentation on the design factors that affect iterative optimizer performance. Our analysis demonstrates that an iterative optimizer performs within 6% of an ideal optimizer and can achieve speedups ranging from 1.03 to 1.55 measured on the SPECint, SPECfp, and mediabench workloads.