As you obviously have profound knowledge on cache matters, maybe you have an opinion on the following as well (if I may bother you again).
Cache density of Banias is really impressive compared to Thoroughbred. In your opinion, could that be due to the different methods the caches are operated in way-ness or has this nothing to do with it (and is just because Intel is couple of months ahead in shrinking SRAM-structures)?
Market Data 
Markets 
AI
