Why Grid Supercomputing Will Just Be Marginally Successful
from the not-quite-what-people-expected dept
Clay Shirky is trying to respond to some of the hype surrounding grid computing by suggesting that it probably has it’s place, but it won’t be “revolutionary” – like some people have predicted. While I actually agree, I’m not sure if I agree with his reasons. He focuses a lot on the fact that most individual users don’t come anywhere near using their PCs to their maximum power, which is absolutely true. However, grid “supercomputing” is not directed at individual users and their computers – and I don’t think anyone has really made that claim. It’s directed at businesses that use vastly more computing resources, centralizing much of them, and allowing them to expand (whether in bursts or over a prolonged period of time) as necessary. As such, I think there is tremendous potential for grid computing – but it’s not likely to change how individuals interact with their PCs very much. In the end, Shirky and I are probably in agreement – and I’m always looking out for topics that have too much hype – but I think he may be incorrectly reading the reaction most people have had to grid computing.
Comments on “Why Grid Supercomputing Will Just Be Marginally Successful”
Why Grid Supercomputing Will Just Be Marginally Su
I think you may both be underestimating the many projects aimed at ‘small’ scale grid computing. Sony is pushing for distributed CPU’s in the next PlayStation incarnation. I believe their partners in this research is Intel(IBM?). The point being, distance between processors is what seems to separate “grid” from “distributed” computing, both are trying to balance the load to achieve break through performance. Finally, _no_ market pushes hardware development like interactive entertainment.
My bet would be on Sony or some other games company solving the distributed/grid computing issue before big business. I suspect it will be the way of the so called ubiquitous computing future.