Great post. For Attempt #3, did you gather the data for branch mispredicts and bad speculation?