Hi,

In principle, the techs try to feed an equal work amount in volunteered resource (CPU time) into the pool, but there are reasons why that can never be completely balanced, amongst very large preference for particular projects and the receiver (the project scientists) not able to take in the data quick enough (the amount in numbers or size), or a single project taking a very large part of the storage. You then have to limit the flow for those. Yesterday you may have seen the note by uplinger that RICE has resolved the data receiving bottleneck, so now they've been throttled up for more units proportionally being put into the shared memory pool... see the effect, where they got immediately 10 more CPU years, of course at the price of less work processed for the others, mostly the biggest.



Just added the percentages in the legend to indicate the current day share from the total!

No the work is largely managed on basis of work units, not time, so it's a juggling act to get the fair amount of CPU time out.

But, your choice is not pointless, because you made the contribution to the project you feel closest to. Eventually, with a large number of people doing that, you see more going to project A than project B.