Ed
04-09-2009, 02:37 PM
Article From: http://www.duelingdiatribes.com/2009/04/onlive-article-draft-do-not-post.html
OnLive, Cloud Computing, and the (not so) Distant Future of Gaming
Recently OnLive has become the buzz word gamers, traders, and tech gurus are talking about, and with good reason. OnLive promises to revolutionize the gaming world if it takes off by eliminating the need for consoles. But what is OnLive and what does it really promise to do?
OnLive was started over nine years ago by Steve Perlman of Quicktime fame.Other prominent employees are Mike McGarvey, a former CEO of Eidos plc and Tom Paquin,
founder of Netscape. How dies it work? On its website, OnLive states
POWERFUL GAMING MADE EASY
Connect to OnLive with your TV, PC or Mac and start a game
Your game runs in a state-of-the-art OnLive game server center
OnLive connects you to game servers through the Internet, instantly sending your controller actions upstream and the results back downstream at blinding fast speeds
Enjoy ultra high-performance gameplay on your TV or entry-level PC or Mac
In essence, OnLive is describing something known as cloud computing. I won't get into describing the architecture of cloud computing here but simply put, someone else hosts a resource that you want and provides it to you, often over the web. OnLive hopes to stream video games and their information from one of its super computers over to consumers. The consumer wouldn't need a game disc or a console. All the consumer needs is a screen to play the game on with any neccesary converter boxes and controllers. Cloud computing itself may be the wave of the future. Cloud computing efficiently manages bandwidth and it is being hailed by many companies as the next paradigm shift in IT. OnLive's website states that one need only a 1.5 mbps internet connection to get regular games and a 5 mbps connetion to get HD games.
All of this sounds great and theoretically, OnLive sounds like a great idea. There are some major caveats though that prevent this technology from advancing as currently projected.
1. Cost: While cost is not an insurmountable issue, there will be immense startup costs and upkeep costs. The closest case studies I can think of are Blizzard's World of Warcraft. Blizzard's upkeep has been about $200 million or so for WoW since 2004. WoW divides people into different server realms. OnLive needs a server per person playing. Think of it like instead of keeping your console at home, you keep it somewhere else and you can access it whenever. You still need a single console to play on though. The theory is that only a small percentage of gamers are on at any one time so OnLive could theoretically buy 10% of all consoles and be able to support a network of almost all gamers, assuming no more than 10% of all gamers are on at a time. Currently, about 12 million Xboxs have been sold, so OnLive would need to purchase about 1.2 million console units. More than likely, the console unit will be a computer rather than an xbox as computers are easier to customize. The price of a decent gaming computer is going to be at least $500 for a base system. The total cost can be highly variable but needless to say, it should cost at least $1 billion for this company to startup and most likely, well more than that. In this market, that is damn near impossible financing to come by. I will add that WoW makes about $800 million profit a year with 12 billion active subscribers and they make over $1 billion in revenue (and its growing) so if OnLive does pan out, it can make this startup money back pronto, especially if upkeep only costs a couple hundred million a year.
2. Lag: If online games have taught me anything, its that lag can make even the most sane person go berserk. Heck, I have friends who play Smash Bro's on an old TV because the lag on their HDTV's is too high. OnLive has two problems with lag. There is the lag that stems from the video encoder, and there is the lag that stems from other sources such as one's TV. This is all additive. OnLive claims a lag time from its encoder of only 1 ms. Thats 1000 fps. This is a few orders of magnitude behind anything we currently have available but it may be possible. Assuming everything works fine, the only lag is 1ms plus whatever lag one's TV or LCD has which in LCDs is about 12 ms while TVs are close to 29ms plus the lag from servers. This is a problem as the just noticeable difference for perceptual latency (at what lag do you notice there is lag) has a low of 16 ms and a mean close to 30 ms. A 1 ms addition is relatively large and brings the TV lag up to 30 and this assumes everything is perfect. More likely, the lag is 5-10 ms minimum and will make quick action games unplayable as those of us blessed with faulty perception who don't notice the lag on our TVs now will notice the lag. Unless the baseline lag in TVs comes down or unless everyone gets an LCD, the added lag from OnLive will condemn the service to failure. Then there is server side lag. A good comparison is again WoW. With WoW, I play at 30 fps or 60 fps depending on my internet connection but the lag is noticeable and around 70 to 100 ms. Even if OnLive gets the lag to consistently less than 100 ms, this is still a huge lag period. As stated, anything in the range of 10-30 ms is noticeable so even without any lag from an encoder, the ability to play quick action games is compromised. There is no feasible workaround for this at this time that isn't expecting the cooperation of Cable companies and of ISPs (discussed below)
3. Gaming culture: OnLive may work well in China where gamers are used to not owning their games or consoles but renting them at an internet cafe. In America, rental consoles and rental game models have not caught on mainstream enough to justify the minimum $1 billion in financing OnLive would need. The best casestudy I can think of was sega channel. It only lasted 4 years and had at best 250,000 subscribers. While these numbers are impressive, recall that for $15 a month, subscribers had access to 50 games. Sega channel also relied heavily on the cooperation of cable companies to work and each installation cost the cable companies $100 and required massive upgrade of infrastructure. OnLive would need to do similar things as ISPs would need to hosts OnLive's servers to combat lag issues. This could decreases lag into the sub 50 ms level but it requires servers and infrastructure at local ISPs. But more importantly, American gamers buy games. We buy consoles. We camp outside Gamestop for that next game or console as a badge of honor. We show it off to friends and customize it and leave our mark. It makes more sense to just rent games from Blockbuster or to buy used, but we buy new. I don't see OnLive changing gaming culture and so I see it as failing.
OnLive has a great proposal and it is a great idea. I fear that the technology is inherently flawed and though there are solutions for it, these solutions require the cooperation of ISPs and/or Cable Companies and none of this will occur anytime soon. The added issues of gaming culture and redtape from console makers could also delay this system's success for years. Give it two years and maybe a decent network will be running but for now, OnLive is just a pipe dream. I do hope I get into their beta though. Sign up here.
As an aside, OnLive could be great for publishers and terrible for console makers and game sellers so perhaps these two lobbies can compete against each other. If I had to choose Activision and EA vs. Microsoft, Sony, Walmart, and Gamestop, I'd choose the latter group every time.
OnLive, Cloud Computing, and the (not so) Distant Future of Gaming
Recently OnLive has become the buzz word gamers, traders, and tech gurus are talking about, and with good reason. OnLive promises to revolutionize the gaming world if it takes off by eliminating the need for consoles. But what is OnLive and what does it really promise to do?
OnLive was started over nine years ago by Steve Perlman of Quicktime fame.Other prominent employees are Mike McGarvey, a former CEO of Eidos plc and Tom Paquin,
founder of Netscape. How dies it work? On its website, OnLive states
POWERFUL GAMING MADE EASY
Connect to OnLive with your TV, PC or Mac and start a game
Your game runs in a state-of-the-art OnLive game server center
OnLive connects you to game servers through the Internet, instantly sending your controller actions upstream and the results back downstream at blinding fast speeds
Enjoy ultra high-performance gameplay on your TV or entry-level PC or Mac
In essence, OnLive is describing something known as cloud computing. I won't get into describing the architecture of cloud computing here but simply put, someone else hosts a resource that you want and provides it to you, often over the web. OnLive hopes to stream video games and their information from one of its super computers over to consumers. The consumer wouldn't need a game disc or a console. All the consumer needs is a screen to play the game on with any neccesary converter boxes and controllers. Cloud computing itself may be the wave of the future. Cloud computing efficiently manages bandwidth and it is being hailed by many companies as the next paradigm shift in IT. OnLive's website states that one need only a 1.5 mbps internet connection to get regular games and a 5 mbps connetion to get HD games.
All of this sounds great and theoretically, OnLive sounds like a great idea. There are some major caveats though that prevent this technology from advancing as currently projected.
1. Cost: While cost is not an insurmountable issue, there will be immense startup costs and upkeep costs. The closest case studies I can think of are Blizzard's World of Warcraft. Blizzard's upkeep has been about $200 million or so for WoW since 2004. WoW divides people into different server realms. OnLive needs a server per person playing. Think of it like instead of keeping your console at home, you keep it somewhere else and you can access it whenever. You still need a single console to play on though. The theory is that only a small percentage of gamers are on at any one time so OnLive could theoretically buy 10% of all consoles and be able to support a network of almost all gamers, assuming no more than 10% of all gamers are on at a time. Currently, about 12 million Xboxs have been sold, so OnLive would need to purchase about 1.2 million console units. More than likely, the console unit will be a computer rather than an xbox as computers are easier to customize. The price of a decent gaming computer is going to be at least $500 for a base system. The total cost can be highly variable but needless to say, it should cost at least $1 billion for this company to startup and most likely, well more than that. In this market, that is damn near impossible financing to come by. I will add that WoW makes about $800 million profit a year with 12 billion active subscribers and they make over $1 billion in revenue (and its growing) so if OnLive does pan out, it can make this startup money back pronto, especially if upkeep only costs a couple hundred million a year.
2. Lag: If online games have taught me anything, its that lag can make even the most sane person go berserk. Heck, I have friends who play Smash Bro's on an old TV because the lag on their HDTV's is too high. OnLive has two problems with lag. There is the lag that stems from the video encoder, and there is the lag that stems from other sources such as one's TV. This is all additive. OnLive claims a lag time from its encoder of only 1 ms. Thats 1000 fps. This is a few orders of magnitude behind anything we currently have available but it may be possible. Assuming everything works fine, the only lag is 1ms plus whatever lag one's TV or LCD has which in LCDs is about 12 ms while TVs are close to 29ms plus the lag from servers. This is a problem as the just noticeable difference for perceptual latency (at what lag do you notice there is lag) has a low of 16 ms and a mean close to 30 ms. A 1 ms addition is relatively large and brings the TV lag up to 30 and this assumes everything is perfect. More likely, the lag is 5-10 ms minimum and will make quick action games unplayable as those of us blessed with faulty perception who don't notice the lag on our TVs now will notice the lag. Unless the baseline lag in TVs comes down or unless everyone gets an LCD, the added lag from OnLive will condemn the service to failure. Then there is server side lag. A good comparison is again WoW. With WoW, I play at 30 fps or 60 fps depending on my internet connection but the lag is noticeable and around 70 to 100 ms. Even if OnLive gets the lag to consistently less than 100 ms, this is still a huge lag period. As stated, anything in the range of 10-30 ms is noticeable so even without any lag from an encoder, the ability to play quick action games is compromised. There is no feasible workaround for this at this time that isn't expecting the cooperation of Cable companies and of ISPs (discussed below)
3. Gaming culture: OnLive may work well in China where gamers are used to not owning their games or consoles but renting them at an internet cafe. In America, rental consoles and rental game models have not caught on mainstream enough to justify the minimum $1 billion in financing OnLive would need. The best casestudy I can think of was sega channel. It only lasted 4 years and had at best 250,000 subscribers. While these numbers are impressive, recall that for $15 a month, subscribers had access to 50 games. Sega channel also relied heavily on the cooperation of cable companies to work and each installation cost the cable companies $100 and required massive upgrade of infrastructure. OnLive would need to do similar things as ISPs would need to hosts OnLive's servers to combat lag issues. This could decreases lag into the sub 50 ms level but it requires servers and infrastructure at local ISPs. But more importantly, American gamers buy games. We buy consoles. We camp outside Gamestop for that next game or console as a badge of honor. We show it off to friends and customize it and leave our mark. It makes more sense to just rent games from Blockbuster or to buy used, but we buy new. I don't see OnLive changing gaming culture and so I see it as failing.
OnLive has a great proposal and it is a great idea. I fear that the technology is inherently flawed and though there are solutions for it, these solutions require the cooperation of ISPs and/or Cable Companies and none of this will occur anytime soon. The added issues of gaming culture and redtape from console makers could also delay this system's success for years. Give it two years and maybe a decent network will be running but for now, OnLive is just a pipe dream. I do hope I get into their beta though. Sign up here.
As an aside, OnLive could be great for publishers and terrible for console makers and game sellers so perhaps these two lobbies can compete against each other. If I had to choose Activision and EA vs. Microsoft, Sony, Walmart, and Gamestop, I'd choose the latter group every time.