Comment on the paper. Provide a short (a few sentences) summary of the focus of the paper or key arguments made. Make sure to specifically address these points:

1. What assumptions were made about computer organization before adding SMT?
2. How does it compare to Wall's claims of ILP limits?
3. What changes were made to add SMT?
4. What performance advantages are claimed? For what workloads?

Your comments do not need to be long at all; I mostly want to make sure that you've thought about the important parts of each paper.