Alex
03-21-2010, 03:16 AM
100 million PS3s is a good number to work with because its covers any potential vindication of Sony's PS3 investment. Lets use Otoy's numbers because they could easily be a down the line vendor for Onlive. According to Otoy 10 of its servers will fit in a 20*20 ft space and that one system or super rack(s) of about 7500 chips would serve 30 thousand HD level clients simultaneously, more standard def clients and a million casual clients.
That means a 10 story building could serve 100 million, better than cycle for cycle, PS3 equivalents. With the figures given by Otoy it would be roughly 6x more power efficient and I am guessing it would cost much less than 1/6 the price of an equivalent number of PS3s. Of course 100 percent of the people would never be on line at one time. It would probably be hard to have a fifth of them online at one time, so divide the cost of the system by 5x and shave off 8 stories from the building. Such a implementation might be had for the price of a new Vegas hotel.
Of course there aren’t 100 million PS3s in the world, that’s 4-5x the actual installed base. Cut that in a third again for what’s in the US and less for Japan or the EU. Of course any one building would only serve about a third of say the US (1k mile radius) so cut that in a third again. So such a building might cost 1/50th the cost of a big Vegas Hotel. About three would be needed to cover the US. That’s 3 for the US, one for the EU and one for Japan. So that effectively replaces PS3 in the world for 1/5 the cost of a Vegas hotel. A mere 100-200 million? No wonder Otoy said it was only 10 cents per high def gaming hour.
Considering Otoy's 10 cents per hour estimate, even if someone played 3 hours a day, every day of the month that's only 9$s a month or the price of an Xbox live subcription- no wonder Onlive thinks it can get away with not offering a bundle subscription. But we can see that the price overhead for serving literally everything (nothing is more demanding than gaming) is small. Small enough to forgo ads and take market share because no other medium is even remotely competitive.
Onlive says it will continually build up the hardware but Otoy sounds like a threshold system especially when its power is used in aggregate where all the cores are used to render one coherent world and the client’s only have a window on it without the endless waste of running full redundant game cores locally. What’s described seems fully capable of real time Hollywood level in game rendering. These might be 20- 40 million dollar installations with 100k chips in them. That’s only 13-14 of the 10 server systems and it would fit in a one story 75ft on a side floor. They aren’t even buildings they are just installations that can go in already available rented commercial space.
So its practically free to destroy Playstation. This is even better than plastic disk toll roads providing publisher free resale so that the publishers that were all duped into being paranoid over a market owed by the public ended up locking themselves out and the public in. Now the public and publishers will be able to lock out the #1 purveyor of these overpriced locks and paranoia driven backwards marketing approachs.
Onlive could offer what they've announced, the limited subscription and demo system. Nothing would stop them from also offering an ad subsidized option and also a bundled subscription, but the bundled subscription that includes everything should certainly not cost more than Gamefly because Onlive is trading its limited library (relative to Game Fly's) for instant higher quality access. They should also offer ad free options- sponsorship mechanisms, obsolete as they are, are what is behind socially useless businesses and institutions.
Skeptics of the actual technology are hard to find now. According to Bioware founder Ray Muzyka, Onlive works. According to Alex St. John only people who don’t understand the technical side think Onlive can work at this point. Muzyka said Onlive works very recently in a Motley Fools article. Remaining skeptics seem to have a vested interest, St. John does because he trying to push a product that would be much less impressive if cloud streaming is available sooner rather than later.
That means a 10 story building could serve 100 million, better than cycle for cycle, PS3 equivalents. With the figures given by Otoy it would be roughly 6x more power efficient and I am guessing it would cost much less than 1/6 the price of an equivalent number of PS3s. Of course 100 percent of the people would never be on line at one time. It would probably be hard to have a fifth of them online at one time, so divide the cost of the system by 5x and shave off 8 stories from the building. Such a implementation might be had for the price of a new Vegas hotel.
Of course there aren’t 100 million PS3s in the world, that’s 4-5x the actual installed base. Cut that in a third again for what’s in the US and less for Japan or the EU. Of course any one building would only serve about a third of say the US (1k mile radius) so cut that in a third again. So such a building might cost 1/50th the cost of a big Vegas Hotel. About three would be needed to cover the US. That’s 3 for the US, one for the EU and one for Japan. So that effectively replaces PS3 in the world for 1/5 the cost of a Vegas hotel. A mere 100-200 million? No wonder Otoy said it was only 10 cents per high def gaming hour.
Considering Otoy's 10 cents per hour estimate, even if someone played 3 hours a day, every day of the month that's only 9$s a month or the price of an Xbox live subcription- no wonder Onlive thinks it can get away with not offering a bundle subscription. But we can see that the price overhead for serving literally everything (nothing is more demanding than gaming) is small. Small enough to forgo ads and take market share because no other medium is even remotely competitive.
Onlive says it will continually build up the hardware but Otoy sounds like a threshold system especially when its power is used in aggregate where all the cores are used to render one coherent world and the client’s only have a window on it without the endless waste of running full redundant game cores locally. What’s described seems fully capable of real time Hollywood level in game rendering. These might be 20- 40 million dollar installations with 100k chips in them. That’s only 13-14 of the 10 server systems and it would fit in a one story 75ft on a side floor. They aren’t even buildings they are just installations that can go in already available rented commercial space.
So its practically free to destroy Playstation. This is even better than plastic disk toll roads providing publisher free resale so that the publishers that were all duped into being paranoid over a market owed by the public ended up locking themselves out and the public in. Now the public and publishers will be able to lock out the #1 purveyor of these overpriced locks and paranoia driven backwards marketing approachs.
Onlive could offer what they've announced, the limited subscription and demo system. Nothing would stop them from also offering an ad subsidized option and also a bundled subscription, but the bundled subscription that includes everything should certainly not cost more than Gamefly because Onlive is trading its limited library (relative to Game Fly's) for instant higher quality access. They should also offer ad free options- sponsorship mechanisms, obsolete as they are, are what is behind socially useless businesses and institutions.
Skeptics of the actual technology are hard to find now. According to Bioware founder Ray Muzyka, Onlive works. According to Alex St. John only people who don’t understand the technical side think Onlive can work at this point. Muzyka said Onlive works very recently in a Motley Fools article. Remaining skeptics seem to have a vested interest, St. John does because he trying to push a product that would be much less impressive if cloud streaming is available sooner rather than later.