One of the greatest privileges of my job is the opportunity to interact with great thinkers such as Jim Gray.
Gray and several colleagues recently published this paper discussing peta-scale data sets in science and the repercussions of such. They are increasingly common, and are most decidedly non-portable -- here's another paper (a great read) by Gray et al that discusses the feasibility of moving large datasets. Based upon this and the observation that datasets are growing faster than bandwidth, that we are going to see a resurgence in data centers -- this time because of the data, not the computation ability. In fact, Gray argues here why it's highly desirable to move the computation to the data and not the other way around.
They then go on to discuss the implications of peta-scale data centers, including an increasing importance of highly scalable databases, metadata, parallelism, and the emergence of a "scientific notebook."
I love that Jim's writing is so readable and accessible by the layperson. Enjoy.
8:36:39 PM  ; ;
|
