I'm participating today and tomorrow in the China-America Networking Symposium (CANS), which this year has a particular focus on grid. It's good to see friends from China, although several are not here because of visa problems. (A familiar, and painful, story.)
I've had the good fortune to visit China several times in recent years. In addition, we have hosted several visitors from China, and I also have some wonderful Chinese students. So I know a little about Chinese grid activities, which include several major deployments, including:
- China National Grid, which links supercomputers and storage systems at eight sites across China;
- China Grid, which is deployed across many Chinese universities; and
- CROWN Grid, which links several sites (while in English, Google translates it pretty well).
All three systems (and others, such as the Shanghai Grid) build on Globus software, which provides a certain degree of commonality. However, both the Pacific Ocean and China are large, and thus it is perhaps not surprising that all three have built higher-level services that are not fully compatible with both equivalent Globus services and each other. However, we continue to meet and talk, and I hope to see convergence in software and interfaces. For example, collaboration with China Grid led to the integration of dynamic service deployment in GT4, and we are working with the CNGrid's Grid Service Markup Language.
The question was raised at CANS of how to enhance China-US cooperation in Grid/eScience, which apart from our Globus connections, are less strong than current EU-US cooperations. The answer is simple, if hard to do: we need more resources to support visits and joint work. The EU is putting a lot of money into this (e.g., EUChinaGrid), and it shows.
For those wanting to learn more about what is going on in China, next year's Grid and Cooperative Computing conference is a good place to start.