Kevin Schofield, General Manager, Microsoft Research
The second session of the day was given by Kevin Schofield, General Manager, Microsoft Research. He has a cool job!
Microsoft Research was modeled after research at Carnegie Mellon University. Researchers pick the problems, independently, across the breadth of computer science.
Microsoft Research is very aggressive about publishing papers (3700 papers in 15 years). Now that it’s a mature organization, it releases hundreds per year.
Academic programs – don’t tell researchers who to work with, and encourage publishing. They are trying to drive forward through research and they also hold workshops, to which a lot of people are invited.
Microsoft Research Mission:
- Advance the State of the art in our chosen areas of computer science
- Making computers easier to use
- Reducing the cost of computing
- Making software developers more productive
- Transfer the resulting technology to Microsoft businesses
- Ensure that Microsoft has a future
· Provide Corporate agility – rapid response to change
- New Technologies
- New competitors
- New Business models
· Reservoir of technology
· Pool of Expertise and smart people
He did a demo of Gigapixel panoramic photography. The zoom was incredible! We started out by looking at a photo of the Seattle skyline and ended up zooming far enough in to see 9 people standing on the observation deck.
Looking at surfaces with Vision.
He also showed us some research that Andy Wilson has done, prototyping The video was cool, but I think that , which was done by a about which I know very little, was better. I had previously seen this demo at .
He also showed us some data from a study that they did using a “Sense-cam” with memory impaired subjects. These people were recorded for 1 day using a sense-cam (that video-recorded everything that they did, and then they were told to watch the sense-cam recording every day for 2 weeks. Not only did their memory of that day improve, but they were able to recall more and more details as time progressed. The surprising result was that they could still remember for 3 months after stopping. This was a small study, but it had some interesting results (not known if they would be repeatable). Microsoft is now working with memory researchers around many different memory impairments to see how this might help people maintain memories.
Microsoft Research addresses challenges for the company. Kevin described some of those challenges as follows:
- Today’s CPU architecture: heat becoming an unmanageable problem
- Power density too much of a problem. That’s why going to multi-core.
- But, no generalized models for highly concurrent software.
- Third problem: code is too big
How to fix:
Build smaller components that are loosely connected. Achieve verifiable composability
Loosely coupled, asynchronous, concurrent, composable, decentralized, resilient systems