> Analyze the problem, divide-and-conquer, then use high level abstractions that make parallelism a responsibility of the underlying implementation. That's a b

> Analyze the problem, divide-and-conquer, then use high level abstractions that make parallelism a responsibility of the underlying implementation. That's a better summary than mine, but I still think it's too little substance for a 60 min talk. If you want a parallel(izable) implementation then divide-and-conquer is a well-known schoolbook approach. The problem as I see it is that it's not easy to build (or understand) algorithms like that. I saw very little in this talk about how to address/solve that, except the obvious: better education. I think it was also a bit irritating that he pretended there was nothing a compiler could do about the accumulation loop he presented as the first example. Isn't it obvious that a smart compiler could (in principle) optimize that loop so that it could run in parallel on multiple cores...?