It’s always exciting when you get invited along to somewhere cool, and this year’s annual MVP gathering is being held next week at Microsoft’s UK Research Centre in the historic university city of Cambridge. The UK has around 200 MVPs and some, including myself will be giving talks and seminars at the event.
I will be giving two talks and a workshop, the talks being on Windows Phone and Windows 8 Tablets and the workshop being a more fun (we hope) Fantasy Microsoft 2021 event with Research staff on hand to help design the next generation of products.
The talks given by Microsoft Research staff will be the most interesting though and I’ll be feeding back here with news and photos from the event, and hopefully some secret tid bits and pictures of cool new prototypes.
The talks I’ll be reporting on (for all I’ll be allowed to report anyway) are…
Programing Life
Our cells are highly sophisticated computational machines, constantly working to keep us alive and free from disease. he software that allows a cell to function is stored inside the cell as DNA, which codes for proteins that ultimately determine the cell’s behaviour. Cells can be reprogrammed by inserting artificially synthesized DNA that can completely transform their function. This ability to program cells could help address some of the many challenges facing humanity. For example, cells could be programmed to convert the sun’s energy into fuels or electricity in a sustainable way, or to fight cancers and viruses more effectively. In spite of this promise, many fundamental challenges still lie ahead, including how to design and implement cell programs, and how to ensure that these programs behave as expected. As cell programs become increasingly complex, computer software is beginning to play a major role in helping to address these challenges.
This demo presents computer software for programming cells. The software can be used to write a program describing the desired behaviour of the cell, and to automatically generate the DNA code that is needed to achieve this behaviour. The software can also be used to program DNA circuits that perform logical computations. Such circuits could eventually be used to monitor the health of a cell and to compute appropriate responses if the cell becomes diseased. In the future, we anticipate that software for programming cells could pave the way for a new wave of biotechnological innovation.
Co-opetition and Game Theory
In many real-life scenarios, we must cooperate with others and form teams to achieve something useful. However, in most of these situations, the reward is achieved by the team as a whole, and we must then negotiate how to split this reward among the team members. These situations have both cooperation and competition elements in them – a phenomenon sometimes referred to as “Coopetition”. Cooperative game theory can help analyse such situations and devise good strategies. I will discuss such solutions, providing many examples:
- How should pirates who must cooperate to find a hidden treasure share it?
- What can block good agreements between service providers in the health industries, and how can subsidies help solve the problem?
- How should we split a bankrupt firm’s funds among its creditors? And how is this related to an old story about a man who had three wives?How we built (arguably) the world’s best Carbon Model
Huge uncertainty in projections of future global CO2 concentrations, and hence huge uncertainty in the future global climate, is just one example of why we need better predictive models of natural and semi-natural ecosystems (other examples include water and food security and pandemics). But the development of such models is currently enormously hampered by a lack of environmental software tools that are intuitive, mutually compatible, and yet capable of scaling up to tackle the largest, most difficult problems. Which is where CEES comes in. Our group tackles real, and really hard, scientific problems, developing whichever scientific tools we think we need along the way. In this presentation and Q&A we’ll explain how we recently developed what we consider to be the most defensible model of the global carbon cycle (think, how carbon cycles among the atmosphere, plants and soil) using several of our new tools including: FetchClimate, a .NET API (together with Silverlight GUI) for intelligent, cloud-based retrieval of complex climate and soil data queries; Filzbach, a .NET library (again, with a Silverlight GUI) for efficient Bayesian parameter estimation on arbitrarily complex models; Dynamic Data Display, a .NET library for interactive visualization of scientific data (think, Excel charts meet Bing Maps meets Gap Minder); and SDS, a .NET API for flexible input, output, and storage of large scientific data sets. With these pieces, we were able, for the first time (and on a laptop!), to produce a simulation model of the global carbon cycle where every sub-model, and every parameter value, has been carefully derived from and tested against appropriate data.
Big Data Systems and Networking
There is an ever-increasing amount of data being produced and correspondingly a need for cheap, easy, large-scale analytics. Cloud storage, cloud computing, and map-reduce have already made “big data” analytics accessible to a much larger audience than before, but there are still many technical challenges left at every level of the hardware and software stack. There are several research projects at MSRC System & Networking aimed at solving these challenges. I will give a broad overview of these projects and more detail on one specific project: reducing the bandwidth usage of map-reduce programs by using static program analysis tools.
These talks should be extremely interesting. In addition there will be talks and seminars from other MVPs, which sadly I won’t be able to get to as I’ll be presenting mine, on subjects including SharePoint, ASP.NET, PowerShell, C#, Behaviour Driver Development and System Centre 2012.
© Mike Halsey (MVP) for gHacks Technology News | Latest Tech News, Software And Tutorials, 2011. | Permalink |
Add to del.icio.us, digg, facebook, reddit, twitter
Post tags: ghacks, microsoft research, mvp
0 comments:
Post a Comment