Nothing about WCG is fast (except your overclocked crunchers 
). It takes a lot of time to compile and make sense of all the data that is run. Frankly, each Work Unit is only a minuscule amount of information (ie, one nanosecond in the folding or stable interactions of a protein). If you want this to turn into a cure however, then consider that there is an intense amount of research that must still take place after the computations. Computer modeling is great stuff, but it is just that, modeling. Its relation to the actual biology varies with the degree of precision specified in the programming model, and the initial observations used to create the model. Not only that, but you have to take into consideration how many variables you want to involve. I can model the stability of Protein X, but in doing so I specify the number of water molecules and other ions and chemicals that are present. Increasing the complexity of the system being tested may make it more accurate (in a limited context), but it also increases the computational time and power needed.
Assuming a reasonable amount of precision and accuracy is obtained however, then you have to do the experiments to verify the biological significance. Granted, there are different projects in WCG, each with a different goal. So, the path may change a little bit. The basic idea is the same though: you have to verify function and interaction of a protein in vivo. Drug companies do similar things on a large scale; they use computation to narrow down the list of chemicals capable of achieving a specific interaction, and then they do the actual testing on the smaller list. This is what WCG does essentially-taking a set of proteins/chemicals/biological variables and seeing what happens in a model system. This establishes a series of predictions that can then be used in further research. It may yield useful data, it may not. So, do we get closer to cure X? Definitely! We've narrowed things down to a manageable size. We just have to continue working on wherever that data takes us.
EDIT: this is mostly in response to Particle, sorry if it seems a little off-topic.
CPU and GPU are both useful, even necessary (the GPU necessity being primarily in amount accomplished per amount of time). You can't go wrong with either...
EDIT 2:
Initial mental response to my post will be disappointment. I apologize. To better see the scope though, your essentially making a cure possible. Having an slightly-less-than-infinite (you know what I mean) list of possible chemicals or whatever to test doesn't make for a good probability of finding any cure in our lifetime. The task of narrowing down the possibilities can't be under appreciated in this context then, as it gives us the ability to conquer the complex tasks of disease research in ways that we couldn't have dreamed of only a decade ago.
Originally Posted by Particle
Reply With Quote
Particle's First Rule of Online Technical Discussion:
Rule 1A:
Rule 2:
Rule 2A:
Rule 3:
Can't seem to find a job any where 
Really rains on my crunching parade, I want to get a 2nd quad up so bad 
Bookmarks