Cloud Gaming, or a video game run on a remote server, is an idea many are excited about. Of course, many have BEEN excited about it for several years now, and it's yet to take off. Many reasons have been cited, consumer acceptance of cloud gaming, lag inherent in running a game offsite rather than locally, etc.

But how exactly are the providers of cloud gaming supposed to make money? This is a question that OnLive, the first consumer facing cloud gaming service, never answered. It is now all but defunct, owned by a consortium that bought it on the cheap in the hopes that Cloud Gaming just might take off yet.

So we have to ask ourselves, what are the costs involved in running a cloud gaming service, and how's it going to sell itself?

First the hardware is needed. Racks and racks of servers need to be purchased to run the games in the first place, and since they need to run triple A video games the racks aren't going to be the cheap, easily replaceable kind of servers that Google and others run on. They'll need proper specs, and with properly spec'd GPUs running alongside CPUs they'll not be an easy thing to cool either.

Another requirement is that they're going to have to be local to the client using them. Ping time adds to the lag time from a users input to what they get on the screen, and this important for highly time sensitive games. And despite optimism by some in the industry for "games built around cloud gaming" there's really no overcoming the speed of light, a cloud game will ALWAYS be slower in this regard than the same locally run one.

And even for less time sensitive, non competitive games lag from input to output can be noticeable and annoying if it's too much, so there's no way to make servers too remote. Which means a provider need almost one usable server slot available per potential user at any time, as there's going to be little tolerance for a "waiting que" for loading up a game.

This means any cloud computing service is going to have to invest in these servers before ever getting paid. They're also going to have to pay rent on wherever they store the servers, maintain them, pay for bandwidth and electricity, and be doubly sure they have as few outages as possible. After all, if an Xbox brakes it's at least a little the Xboxes fault even if it's mostly Microsoft's, and the Red Ring of Death cost Microsoft billions, though much of that price comes from the customer paying for the hardware directly. Still, a huge server infrastructure is already more complicated than a single box, and if it breaks it's entirely the services fault; and it will inevitably be down. Even such giants as Google, Amazon, and MS have services go down once in a while.

And then we come to the question of payment. Game's can't cost anymore than they do on any other platform, because inevitably customers aren't going to pay it. Look up "price equilibrium" for a basic explanation of why this will always essentially hold true. A monthly service charge to customers also seems a risky bet at best, especially if customers lose access to all "their" games the month they stop paying. People are already unhappy with Xbox Live Gold as it is, losing access to games altogether will probably be almost completely unpalatable. As for charging an upfront fee, this degrades much of the benefit of Cloud Gaming to most consumers, as it leaves little differentiation between The Cloud and Console. The idea that someone might regularly want to play an involved, hardcore game on their Ipad has flown long ago.

Which leaves charging a portion of each game sale to make the cost indirect, as is the case with current consoles. Hardly an endearing prospect to developers, many of whom have already gone independent and seem to love platforms with little to no royalties such as digital distributed games to PC's and mobile platforms.

And Cloud Gaming providers will be making a huge bet that this will be enough revenue to turn a profit. While consoles certainly have large development costs customers largely defray, or even pay for the hardware themselves (eventually to the tune  of a profit for the console maker!) as well as pay for storage, bandwidth, and electricity.

Any Cloud Gaming provider will make a huge outlay on any initial offering of the service regardless of how many people come. And their profits, or revenue at all, will rely on the constant purchase of new games. Every customer that stops purchasing games will become a liability, as the potential for them to play games they already own is always there, and servers need to be up for them to do so. They'll be a double liability if they keep playing old games while not purchasing new ones, as there won't be a way to cheat and get out of not having a server for them to play on.

And none of this adresses that ever present threat in the customers mind of being cut off from the very games they've already "bought", among many other concerns. Clearly, just breaking even on Cloud Gaming is a large proposition, let alone making a profit. It remains to be seen if anyone can actually do it.