I attended a talk on the Intel's experimental 80-core processor given by Tim Mattson at SC'08 this week. Dr. Russell Winder captures one aspect of the talk nicely in his blog entry. It was fun to see Dr. Mattson's energy and enthusiam about some of the concepts brought out by the chip, one of the main ones being no cache. Another being the small instruction set, allowing a certain freedom when programming the cores (but also many constraints, such as the inability to code inner loops).