PDA

View Full Version : Need help assigning shares



jasong
07-28-2006, 04:02 PM
My share allocation at the moment is pretty random. I have the following projects, with the following shares:

uFluids 300 25%
LHC@Home 100 8.33%
Predictor@Home 100 8.33%
BOINC SIMAP 50 4.17%
Rosetta@Home 400 33.33%
Einstein@Home 200 16.67%
Quantum Monte Carlo@Home 50

My thinking is that QMC, which studies the mathematical modeling of subatomic systems, assuming it only has a small number of users, might be the one that, relative to other projects, needs the most computing power.

The third to the fifth from the top were joined on a whim, I don't even pretend to understand the differences between the three.

Einstein@Home is, in my opinion, on the cutting edge of science. Even if we don't discover gravity waves, we've still gained a bit of knowledge from our attempts.

I'm really not sure about uFluids, I have no idea how much power they do, or don't, need, but I think it's a very cool project.

LHC is very obviously getting results and constantly tweaking their designs, although they are very often out of work. I'm not sure if this is because they have more than enough users or some other reason, like having to frequently analyze results.

If anyone has an opinion of what projects are in the biggest need of power, from an available power standpoint, as opposed to a Free-DC rankings standpoint, please post.

Thanks in advance.

Biggles
07-30-2006, 12:15 PM
As always, each to their own.

In my opinion, Rosetta is the one that can best use the power. They have the funding and the resources to do anything with their results. They also make Predictor completely redundant. Predictor does the exact same thing as Rosetta, but not as well and with less input from the scientists and with no admin contact. Predictor treats its users badly and we don't know right now whether or not it's actually doing any useful work.

I like LHC a lot. I'm not sure how it can really do that much useful stuff with the scientific results, but you don't spend billions and billions just out of a small sense of curiosity. Because it doesn't really have that much work to do, I have it on a high resource share so that I do as much as possible when it does have work.

Einstein, like LHC, I'm unsure of the actual worth of knowing the scientific results. I do run it from time to time when the whim takes me. I don't think it's as likely to help humanity as Rosetta is.

QMC and SIMAP, I think do similar work to Rosetta. Not exactly the same, but certainly something to do with Protein prediction or folding, or eating or something. I run both of them on occasion. Whilst I don't know if they are scientifically "better" than Rosetta or more useful to science, I don't think they have the same resources to really make something of the results.

I like uFluids as well, because I also think it a cool project. But at the same time, the project is partly work being done for the admin's thesis. It's less likely to change the world.

Run whatever makes you happiest, whatever you think is the coolest. In my opinion, the most worthy project scientifically is the World Community Grid and Rosetta, followed by Climate Prediction. Mainly because curing cancer or whatever other diseases is a good thing and because global warming IS happening. But I also don't run them the most, because I'm more interested in niche areas.

You could argue that SETI is the most important, because contacting aliens would probably be the biggest thing to happen to humanity since Jesus, but there are also very small odds of success and dubiety over whether or not it would work. But an awful lot of people get their kicks from looking for ET, so who am I to argue? Finding a bigger prime number doesn't really benefit humanity, but it excites a lot of people and I've done it when the whim took me.