…buy another server. 
You find a server for $15, and I’m sure they’ll get right on that. 
How much for a server that could make the SDMB run smoothly?
Welcome aboard the boards, OneCentStamp. (Might be a bit like one of those church roof fundraising displays. As OneCentStamp’s transaction goes through, another lick of paint is added on the loooong way to the top).
Thank you for your contribution(s) to the community, financially and otherwise.
TubaDiva
You have made other contributions, yes?
Mostly inane questions and unconstructive wisecracks. 
Perfect! Welcome aboard!
Hey I need to mail this letter, can I borrow you? Do you have self-sticky glue on your butt?
You’ll need 38 of his friends too.
There are actually two questions here. The first question, easily answered (in principle, at least) is “Given the board’s current level of usage, what kind of server would be needed for smooth running?”. There is some such server, and that server has some price, and that price is equal to some number of subscriptions.
The more complicated question, however, is “What kind of server would be needed such that the board would run smoothly after that server were installed?”. This may look like a simple rephrasing of the first question, but it’s not. At the moment, usage of the SDMB is largely regulated by the level of impatience among the users with the server. In other words, there is some threshhold level of performance above which will be tolerated, and below which will not be tolerated. If we get a better server, then we can accomodate more users at that threshhold level. So more people will join and will post more until the server is again burdened to the threshhold level of performance, at which point people stop joining or posting.
The net effect of this is that, no matter how good a server we have, the performance will always be just above the threshhold of tolerability. The only difference between a better server and a worse one is that the better server allows a greater number of people to enjoy that threshhold level of performance.
This will not continue forever, of course: At some point, we would eventually reach a level where we’ve saturated the market, and all of the users who would be interested in our site are on it. At that point, then increases in capacity would not result in increases in usage, and we could actually see stable increases in performance beyond the threshhold level. But it’s very difficult to estimate just how large the potential market is, and it’s probably huge (orders of magnitude larger than the current population of the board). So for the time being, it looks likely that we cannot improve the performance of the board by upgrading the servers, no matter how much money we spend on them.
Ummm, Chronos. The little winky smiley thing at the end of OP means he was just fooling around.
Whoa, be careful when asking the SDSAB a question.
Well, OneCentStamp winked, but Lakai didn’t. And while the notion of a single subscription paying for a new server is clearly meant in jest, the question of how much, total, a new server would cost (and therefore how much would have to be raised from subscriptions to afford one) is a question which often does get asked in all seriousness around here.
Anaamika, look not to the SDSAB for advice, for they will tell you both yes and no.
You mean I have to lick 39 butts? Better damn well be the self-sticking kind.
My butt is definitely of the self-kissing variety; I see no reason why it couldn’t take itself halfway to second base.
This assumes an infinite amount of potential demand, which is patently ridiculous. This is a popular website, to be sure, but it’s not infinitely popular; there is some upper limit to the number of people who will be reading it at any one time. The question is really “How much money is the Reader willing to spend to ensure that the demands upon the server are met?” And the answer is, as always, “Not much.”
So, we deal with it.
And thus, the hare will never beat the tortise.
But seriously, this is a common problem in game theory, be it in economics, infomatics, or evolutionary zoology. By satisfying existing demand, you create the expectation for demand to be met, which increases demand to be satisfied, ad nausum. As Chronos states, there is an upper bound–ultimately, access by every person with an Internet connection and a desire to fight/be illuminated from ignorance, though in practice much lower–but the issue is at what point it is cost effective i.e. will bring in more or lose less revenue versus cost to enhance performance.
What kind of servers are you guys using, anyway? Some of the throughput and access problems might be alieviated by reconfiguring the system rather than adding more capability. It’s a matter of figuring out what is the worst bottleneck, and expanding it…so that something else can become the bottleneck. :smack:
And I don’t care how much it saves, I’m not licking anyone’s butt. :dubious:
Stranger
[Life Cereal Voice] Let’s give it to 'Mika! She’ll lick anything. [Life Cereal Voice]
Eeew. I think not.
You do have scruples. Updating my list now…
“No butts… and no old wrinkled men’s butts over 60.”
And they are subtle and quick to anger.