LHC@home

I distinctly remember the 50th challenge being described as an attempt to get 50,000 people to join in 50 days. Now it's described as The goal for the challenge is to get many thousands of people to join LHC@home, and to run for at least 50 days

and we get comments like this from Markku

>>I admit that maybe we allowed too many users to join.

Does that mean the T-shirts and coffee table books have been abandoned too?



Giskard - the first telepathic robot.

> I distinctly remember the 50th challenge being described as an attempt to get
> 50,000 people to join in 50 days. Now it's described as The goal for the
> challenge is to get many thousands of people to join LHC@home, and to run for
> at least 50 days
>
> and we get comments like this from Markku
>
> >>I admit that maybe we allowed too many users to join.

Yes, as I also said: we underestimated computing power of "mere" 5000 users. If we at some point get a really big study to do then we will be likely to raise user limit again.

In other words, we already passed this challenge, but with less users than previously thought:) It's more like 5000 users in 5 days... Not exactly true but we did get about 1000 users per day when the user limits were raised.

> Does that mean the T-shirts and coffee table books have been abandoned too?

Not necessarily:)

Markku Degerholm
LHC@home Admin

> Let�s look at this project as if it were a business.
>
> The supercomputer that you've built, as powerful as it may be, is currently
> consuming untold KWh of electricity and thousands of man-hours of
> administrative and maintenance costs, all while idling away the days waiting
> for work. It has become apparent that the device was overbuilt for its
> intended purpose and responsible business practices determine that these
> expenditures constitute an extremely poor investment of company resources. I
> wouldn't be surprised if the shareholders demand that the Board of Directors
> start selling off some of the assets in order to remain profitable!
>
:)

On the other hand, we do have these supercomputing clusters downstairs. They require big MW-class transformers just to supply the electricity, and lots of money, knowledge and administration to keep them running.

Now the 2 old PC servers running this BOINC project produce about the same computing power, but require much less power and operate on (almost) free manpower.

So it's not that bad, after all:)

Disclaimer: LHC@Home does not compete with LCG. They are two separate solutions for two separate problems.


Markku Degerholm
LHC@home Admin

>
> The point that you missed is that these costs still exist, they are being
> absorbed by the shareholders.
>

Yes, now I get your point (long day, long wires) :)

Management will see to it that this supercomputer will actually be used for something useful. It just takes some time to make staff learn how to use it properly.



Markku Degerholm
LHC@home Admin