It happens to me every day, so I’m used to meeting people that are much smarter than me. But still today was a very special day because I had the honor of hosting Andrew Herbert, managing director of Microsoft research in Cambridge and Craig Mundie, chief research and strategy officer of Microsoft Corp.
Andrew, a distinguished engineer himself talks passionate about how research drives the evolution of IT in existing areas but also in new areas like machine learning e.g. computer vision, search, online gaming algorithms. Especially those areas of research that make it into an actual product and have impact on day-to-day live. His quote of the day that summed it all op was “users wanted a cheaper 3270 terminal, an engineer invented the mouse, and some users got exited”.
Craig talked about the pendulum and the balance between centralized and local computing, about Moores law, and how the processor temperature challenge is going to force the race from faster processors toward parallelism fuelled by the race of Multicore processors. Parallelism on itself is not new, but the expansion to any computational areas is, from desktop to cell phones to game consoles to watches. The massive computational power at the edge will have enough idle time to spend on speculative execution (working on those things you might want to do next) and as such maybe even become your proactive human assistant. A pretty strong argument supporting the Software + Services evolution happening now.
I’ll spare you what I presented after these 2 eloquent execs, in the immortal words of Waynes World, I’m not worthy