Reese

OTOY Developing Server-Side 3D Rendering Technology

Forget upgrading your computer every 2 years to play the newest games. I think this is where PC gaming, hell maybe all gaming is going in the next 15-20+ years. Instead of buying a disk, installing gigs and gigs worth of data on your harddrive. All you do is load up your Internet browser and use plugins that you already have (flash, Java, ActiveX) and you have graphics more realistic than the PS3 or Xbox360 with no installing. Everything is steamed to you.

Basically what's happening is, The software, all the models, textures and everything is rendered on their supercomputers, it's turned into a video and streamed to you. It's PERFECT! almost..

Pros:
- Excellent graphics
- No compatibility issues, All you need to do is be stream high quality video.
- Publisher retains full control of the product. You can't break open the game and steal the models or textures, you can't crack the game because it doesn't even need a CD key, just a login and password.

Cons:
- lag - the high quality stream is about 250kbps
- LAG Ever tried playing warcraft while streaming 250kbps?
- LAGG!!! Not only are you streaming 250kbps and trying to send other data to control the stream, since EVERYTHING is server side, you won't see a response to your movement until you upload the command(up , down, left, right), it's rendered and streamed back to you. At least when you're lagging in WOW you can run around and when it catches up, you don't even notice because the server corrects the de-sync. With this, a lag spike pretty much disconnects you.
-Publisher retains full control of the product. You can't play the game offline. When the publisher decides to quit supporting the game, too bad you can't play it anymore.

This is a VERY early technology. Our current internet infrastructure just can't handle 100s/1000s of games like this, it's not until everything is fiber optic that we're gonna going to see this mainstream. The current applications are a perfect launching ground. They're virtual worlds created by one of the founders of myspace. I say they're perfect because there's no real action that needs to take place, It doesn't matter so much if there's a few MS lag between hitting the up arrow and seeing the character move, it doesn't matter if a lag spike hits because there's nothing to lose..

I'll probably have grandkids before this becomes a reality(I don't even have a GF now if you're wondering how long that will be.) But hell, I can't wait.( For awesome real time gaming that looks better than the transformers movie..) not to have grandkids...

__________________
“It is better to be rich and healthy than poor and sick” - Dave Barry

ironpham

This would be fantastic for the work place (for me anyways).

__________________
SIGNATURE.

Baraka_Guru

It's like gaming....televised! This is a cool concept, and would make sense on a number of levels. Data bandwidth is one of the greatest technological challenges of our time.

__________________
Knowing that death is certain and that the time of death is uncertain, what's the most important thing?
—Bhikkhuni Pema Chödrön

Humankind cannot bear very much reality.
—From "Burnt Norton," Four Quartets (1936), T. S. Eliot

Redjake

I think this is awesome

perhaps they could use the servers to render the draw distance only - arguably the most FPS-ruining aspect of games. that way, you aren't streaming the entire game and the bandwidth would be a lot less.

Nisses

I'm going with something like: within 40 years. If you're saying: in the next 200 years, who knows what can happen...


The whole reason why we have the current infrastructure is because of the sheer computing-power you need on the server-side combined with the enormous MASS of data the server has to upload. (think about it, it needs to send a complete and highly detailed stream of images for every computer connected to it, after having rendered them all from each POV)

Somehow I doubt anybody will ever bother to implement any WoW-like MMO ever, simply because of the financial cost.
That combined with the fact that all their players would have to have a very very big internet-connection to even play it.

Sure, you're not paying for extra hardware anymore, but your internet costs will go through the roof.

Somehow, I don't really see any of those pro's as worth it:

Pros:
- Excellent graphics --- for an MMO, I doubt it.
- No compatibility issues, All you need to do is be stream high quality video. -- exactly so, limiting your playerbase severely.
- Publisher retains full control of the product. You can't break open the game and steal the models or textures, you can't crack the game because it doesn't even need a CD key, just a login and password. -- even more keylogging spyware on the market.

__________________
Moderation should be moderately moderated.

Baraka_Guru

Okay, 40 years ago, the best option for a minicomputer for the average person was Data General's SuperNOVA, a 3.3mhz, 16-bit machine that cost over $8,000. And 20 years ago, IBM released their beefy PS/2 286 machine, with 16MB of RAM and 3.5" floppy drive (finally! Small floppies!). And just 10 years ago, Intel released the smaller, faster, cooler 333 MHz Pentium II processor! It was this kind of innovation that made them a world leader....

You think it will take 40 years to have adequate delivery of server-side rendering? You've got to be kidding. This sounds like the kind of thing that people must have said when they were trying to figure out how to broadcast films (moving pictures) into peoples homes via what would come to be known as the television.

Consider this: Gaming and Internet entertainment are going the way of the film and television industry. It's where the money is. It's just a matter of time before something like this will happen. But 40 years? I'm not so sure about that.

__________________
Knowing that death is certain and that the time of death is uncertain, what's the most important thing?
—Bhikkhuni Pema Chödrön

Humankind cannot bear very much reality.
—From "Burnt Norton," Four Quartets (1936), T. S. Eliot

Nisses

Baraka: the reason they managed to become leader, is because of the fact that they opened up a huge and gigantic market in the home-user.

“I think there is a world market for maybe five computers.” – Thomas Watson, chairman of IBM, 1943
His opinion was based upon the huge computers as they were then. And frankly, if you adjust the number somewhat, those types of clusters still don't have a big market.

Who in their right mind would set up a center with such an incredible need for power as what they are suggesting here? It would take years and years to pay off, with rising energy costs as they are, you might not ever break even.

And all this for just ONE application?

This is not a question of possibility or theoretical feasibility, but more a question of WHY would you do this?

A game does not have an endless lifespan. And the next game will require more performant processors, more GPU power and more datatraffic, making your hardware only minimally suited for 'recycling'.

__________________
Moderation should be moderately moderated.

Jinn

I actually researched this during my CS undergrad in our Advanced Computer Graphics class - it seemed feasible within our fiber LAN at the University, but the rendering power was severely limited.

I'm not sure what company would be willing to house a render farm just to support a fan base for a given game. The cost to the company is much greater than the cost to the customer divided over every one of their computer.

Furthermore, this render farm would only support the one game that the game designer made, rather than the hundreds of games you can run on your home PC with one video card.

__________________
"I'm typing on a computer of science, which is being sent by science wires to a little science server where you can access it. I'm not typing on a computer of philosophy or religion or whatever other thing you think can be used to understand the universe because they're a poor substitute in the role of understanding the universe which exists independent from ourselves." - Willravel

Baraka_Guru

Those who can find a market, naturally. Server-side gaming would open up an already large gaming market. The barrier to gaming for many people, especially with top-of-the-line games, is the hardware requirements. If you could eliminate this need for your end user, you suddenly have fewer barriers of entry for those who'd otherwise spend money on it. Personally, I'd pay as much as $50 to $100 a month (more, if I could see the benefits) for access to a gaming environment if it didn't require me to have more than a computer that can run high-quality video and high-bandwidth Internet transfers. And I don't think I'm alone on that. There are likely millions who think the same. Now tell me, how expensive do you think it is to maintain a cable network for a large cable company? How much do they charge their end users for access? I'm telling you, this is akin to what the television industry went through. And think of films. All of their content and "magic" is "server-side." They don't require each of their end user to find a Brad Pitt and a set and have him do the actions. The end user for the film industry does not have to do the work of a production company. Why not the same for gaming? If technology isn't a barrier, then neither should marketability be.

Do you realize how much energy is being used by individual machines to do the work of what a server could do more efficiently? I don't doubt there is a potential market for this.

Why not? Spiderman 3 had a $258 million budget, and that was just ONE film. The key is to make enough money on subscriptions, etc. (or, in the film's case, movie tickets, DVDs, etc).

Games aren't meant to have an endless lifespan, as we know. Each game would need to be approached like a business. Effective business models account for all the factors that need to be taken into account to ensure a profit is made. When a studio approves a film, I'm sure they think this way. After all, $258 million is $258 million.

To run something like Google, it is estimated that 6,000 processors and 12,000 IDE disks are employed. Now tell me, what in hell makes it feasible for a company like Google to invest in that kind of hardware when they don't charge a cent for the end user to access their most popular feature, the search engine? The answer is quite simple, really, it's the monetization of information. If you can monetize something as vague as "information." Why could you not monetize the delivery of server-side gaming? Take both the television industry and Google into account when you consider how long it would take to find a market for this kind of thing.

It isn't the question of number of years; it's a question of number of brilliant ideas. It's just a matter of time. But, again, I don't think something like this would need to take 40 years.

__________________
Knowing that death is certain and that the time of death is uncertain, what's the most important thing?
—Bhikkhuni Pema Chödrön

Humankind cannot bear very much reality.
—From "Burnt Norton," Four Quartets (1936), T. S. Eliot

Reese

Nielsen's Law of Bandwidth says there's a 50% increase in data transfer every year and his empirical data starting from 1983 supports this. Assuming it remains this way for 20 more years, That's going to be HUGE data transfer speeds.
Nielsen's Law of Internet Bandwidth (Alertbox Apr. 1998)

Moore's law has been right also when it said the number of transistors per integrated circuit is doubling every 24 months. The thing is, Moore's law is talking about inexpensive, aka home user machines. Imagine if they could use more expensive materials to create super computers capable of streaming the entire computer experience to 1000s of users at a time. They could then recycle their precious materials and reuse them. The only downloading that would exist would be the stream of 250kbps. Everything would be server side. All programs, all information stored on the server. You could be listening to your favorite music and playing an ultra realistic game with nothing more than a monitor and keyboard. Your bandwidth never goes over 250kbps. No worries about viruses or spyware because the server takes care of it. Big Brother would love it. It's almost scary.

250kbps sounds like a lot but it's not. A lot of people are getting 250kbps right now. Assuming we get anything close to 50% increase in bandwidth per year, That's almost enough to stream 250kbps to every active computer with our current infrastructure. I doubt this will happen in the next 100 years, but if the entire internet was centralized with huge clusters of supercomputer streaming out nothing but video and accepting no data except input commands, It'd make data transfer and management much more efficient.

I had a longer post but I accidentally clicked a bookmark that opened in this tab.. grr..

I'm definitely going to go play in Liveplace when it launches.
LivePlace To Launch Photo-Realistic Virtual World Rendered In The Cloud

__________________
“It is better to be rich and healthy than poor and sick” - Dave Barry

Cynthetiq

I'd also state that hardware is the loss leader and the money is in the licensing of the software to the platform.

Thus, MS, Sony, Nintendo make bank on the licensing fees for every game made and sold for their platform. While it will be attractive to the developers because they only have to develop for 1 platform if there happens to still be something like a wii, then Nintendo will make zero licensing fees to allow it to run on their machine. Not likely to happen in todays marketplace.

As far as the technology is concerned, that is the realm of why the PC games make the leaps and bounds that they do in comparison to the consoles. The PC hardware is still changing and evolving. Consoles stay put for iterations at a time.

I don't see the add value aside from that it's really cool.

__________________
I don't care if you are black, white, purple, green, Chinese, Japanese, Korean, hippie, cop, bum, admin, user, English, Irish, French, Catholic, Protestant, Jewish, Buddhist, Muslim, indian, cowboy, tall, short, fat, skinny, emo, punk, mod, rocker, straight, gay, lesbian, jock, nerd, geek, Democrat, Republican, Libertarian, Independent, driver, pedestrian, or bicyclist, either you're an asshole or you're not.

Jinn

The issue of downtime or malicious DDOS comes up too. People complained for months about having to validate through Valve every time they played a SINGLE-PLAYER game, so I can't imagine the ire of people complaining that they can't play a game they paid for because the render farm is down or unreachable.

Also, what about the substantial market of people who do not play online games or those who do not have "always on" Internet?

__________________
"I'm typing on a computer of science, which is being sent by science wires to a little science server where you can access it. I'm not typing on a computer of philosophy or religion or whatever other thing you think can be used to understand the universe because they're a poor substitute in the role of understanding the universe which exists independent from ourselves." - Willravel

Cynthetiq

yes i was going to add that but figured by the time that this technology was more mature and bandwidth larger, your gamers would all be online and the smallest percentage would be offline.

I prefer offline to online myself.

__________________
I don't care if you are black, white, purple, green, Chinese, Japanese, Korean, hippie, cop, bum, admin, user, English, Irish, French, Catholic, Protestant, Jewish, Buddhist, Muslim, indian, cowboy, tall, short, fat, skinny, emo, punk, mod, rocker, straight, gay, lesbian, jock, nerd, geek, Democrat, Republican, Libertarian, Independent, driver, pedestrian, or bicyclist, either you're an asshole or you're not.

CyCo PL

This would be a lot more feasible once the infrastructure of the internet is improved and everybody is pulling really high speed downloads... Cable/DSL isn't the end.

Nisses

Baraka:

I'll agree that if you can find a market for it, that's the first step.

But you'll need to set your pricing within range of the market.

And as far as the energy consumption goes: you'll still need both the server AND the pcs running, so somehow I doubt you'll be saving much on power.

As far as Google goes: they use their hardware to the fullest, and to the maximum of the lifetime.

Maintaining your park will be alot less costly for your gaming-cluster than it would be for Google, because all your stuff would still be new. What would be important is the cost for the new material.

However, I've noticed that you're talking more along general lines rather than practical limitations, so I'll reply to the general question:

Why wouldn't you be able to monetize server-side gaming?
There's no reason, you would definitly be able to: you're providing a service and a convenience for customers, so that's a no-brainer.

My point is: There's no way I can see that actually makes it financially viable. Bringing me back to: Why would anybody (entrepreneur-wise) want to do it?

Maybe I'm just a pessimist. I guess time will tell.

__________________
Moderation should be moderately moderated.

Jinn

I see a market for server-side rendering in places like gaming internet cafes where they could have a render farm in the back and a lot of cheap thin clients in the cafe itself. The hardware cost of replacing 20-30 high end desktops has caused a few gaming cafes around here to shut down. In this situation the latency is guaranteed to be peak, as it's LAN speed over a guaranteed distance and number of hops.

I could also see a market in the high-end home market. You have some sort of rendering server in your house, garage, basement, etc, and a lot of high bandwidth thin-clients around the house. That way you could play a graphically intensive game on your table, your microwave, tablet PC, etc. without the huge rendering and heat overhead for all of those devices. In this situation the latency is again guaranteed to be peak, as it is LAN speed over a guaranteed distance and number of hops.

I just don't see Internet-based server-rendering as feasible or desired, not the least being that I remember our bandwidth being MUCH higher than 250KB when we toyed with it in University. They must be using some sort of compression algorithm, which defeats a lot of the gain by forcing a compress/decompress step ON EVERY FRAME - 60 times a second, at least..

__________________
"I'm typing on a computer of science, which is being sent by science wires to a little science server where you can access it. I'm not typing on a computer of philosophy or religion or whatever other thing you think can be used to understand the universe because they're a poor substitute in the role of understanding the universe which exists independent from ourselves." - Willravel

Reese

Valve had a major problem. At the time they begin introducing the online authorization not even 50% of US households had always on internet, hell some had no internet at all - Me included. I had to go to a friends house to set up my game and download that first patch.

This isn't going to just be the biggest thing ever overnight. It's starting out small, almost un-noticed. It's starting with a simple "myspace" world where you just hang out. It'll probably be subscription based with in-world advertisements to support the currently very high costs. Who knows how cheap or expensive it'll be finance it. There's a lot riding on the success of this first step into server side rendering. If Liveplace does half as well as something stupid like Myspace, There will be copy cats. The technology will be perfected and it'll lead to all kinds of different interactive worlds.

Heck, There may not even be different servers. The server rendering that city could probably easily render a Fantasy world at the same time. The only difference is which models are used. Just give everyone or at least certain developers the ability to create a wide range of triggers(if, then and else commands) and I guarantee you'll see games within the game.

__________________
“It is better to be rich and healthy than poor and sick” - Dave Barry

UKking

Very interesting concept indeed, but man alive, the network capabilities to run that would have to be absolutely immense. Imagine how pissed you'd be if your network went down during this. The number of factors affecting the speed of your connection is large, and it'd take an enormous amount of bandwidth to comfortably be able to overcome these. I mean, my Netgear wireless thing gets shirty when I move it a quarter of an inch.

Never buy Netgear.

biznatch

Well this is a pretty big step forward in gaming IMO. TRIBES-like multiplayer online FPS in your browser. for free.

__________________
Check it out: The Open Source/Freeware/Gratis Software Thread

RaviNara

There is , of course, a very serious problem with serverside games. Months or years after I buy the game, the company may well be out of business or no longer supporting that game. What am I supposed to do after a game that I paid for in full is now unusable to me?

People will no longer own the game, they will be renting it. And there won't even be an "rent to own" option, if everything MUST be done by the server then you'll be left to the whims of the company.

"Oh, turns out that this game is no longer profitable to keep running, sorry about your time and money but it's gone now. Too bad!"

Count me out. I'll pirate long before I agree to pay to use someone else's game and then have it taken away from me.

Martian

This seems unlikely to me, for the simple fact that nearly everyone else in the digital world is moving away from monolithic design principles, particularly when it comes to processing. It turns out that a lot of small processors are almost always more efficient than one big one.

I could see online data storage. Something like Steam taken up a notch. But to move the actual processing and rendering away from the end user just seems inefficient and unnecessary.

__________________
I wake up in the morning more tired than before I slept
I get through cryin' and I'm sadder than before I wept
I get through thinkin' now, and the thoughts have left my head
I get through speakin' and I can't remember, not a word that I said

- Ben Harper, Show Me A Little Shame

biznatch

What happens nowadays? Say, in Counter Strike Source. Is all the physical data (shots fired, grenade trajectory, player location, speed, direction and position) processed on the server, and sent back to us? Or does our computer also process it?
Does our computer only render it?
I'm not sure about server side rendering, regardless. Considering one needs at least 30 fps for decent gaming, each frame at high resolution (and probably higher resolutions still as years go by), the server would need to render hundreds of frames, from many different points of view, and send them out. It's quite a large amount of processing, and quite a huge amount of bandwidth. Maybe if our internet connections dramatically speed up.

I'm wondering, maybe they could improve flash video to the point where streaming the rendering of the game to the user's computer is a possibility?

__________________
Check it out: The Open Source/Freeware/Gratis Software Thread

UKking

Your computer processes all that stuff. That's why, if your connection isn't great, you'll be able to carry on running and shooting for several seconds, apparently kicking ass because nobody's shooting you back, and then all of a sudden you'll seem to teleport to somewhere else - your PC hasn't received data from the server updating the battlefield, and when it receives it, it has to change where it thinks you are in order to fit in, if you see what I mean.

It happens a lot on Pro Evo 2008, because their servers are shit and they're often pretty laggy. Your team'll have the ball and you'll be running towards goal unopposed for a few seconds, but then the connection updates and you find that the other team has the ball at the other end of the pitch.

biznatch

What does the server process, in terms of "physics" ? Don't they do some processing? If not, what's to prevent a player from modifying his client somehow and cheating?

__________________
Check it out: The Open Source/Freeware/Gratis Software Thread

UKking

Well, Valve in particular uses Valve Anti-Cheat in its games to prevent it, and there are other systems that prevent it such as PunkBuster. You're quite right that there's no way to stop people from modifying the game on their PCs, however, there are ways to tell if someone has done that to give themselves an unfair advantage online. It's not foolproof, but VAC in particular has been very effective, partly because it's adept at detecting cheaters, and partly because if a game needs activation on Steam (Valve's gaming service), as Counter-Strike does for instance, then the threat of being banned from playing is often enough to stop players from trying to cheat. In Steam's early days, Valve were very aggressive in banning cheaters and people who were illegally downloading games, and earned a lot of hatred this way - however, that policy seems to have paid off because people quickly realised that Valve's ruthless approach to cheating and piracy wasn't worth trying to fight (plus the fact that Valve's own games, Half-Life 2 and Counter-Strike Source, needed Steam to run, and the popularity of those games essentially ensured that Steam was here to stay).

Anyway, the talk of Steam and prevention-before-the-cure is getting off the point somewhat.

Basically, here's what happens: The server does no processing in terms of game content; all it does is provide the common hub for game data to be exchanged between players. The player's PC has a copy of the game which is verified by whichever anti-cheat system is being used, and what that means is that the server believes that your version contains the same code as everyone else's. So gravity is the same for you, the number of bullets to a gun is the same, the guns' power is the same, you don't have noclip enabled or anything like that... the playing field is level. These things are regulated by the server - that is to say, everyone's game has to match up with what the server will accept; not only this but the server admins can change the rules in-game, and often these are put to the vote, so you might to say whether you want the change the map, or whether you want low gravity, or a knives-only round, or any number of other changes.

If a server doesn't use an anti-cheat system then, AFAIK, there's nothing to stop you from doing what you like. Unless the other players get pissed off with you doing so well and have you banned. Most players are honest, and they'll get very annoyed very quickly with cheaters.

The main advantage for getting the server to process game content is that it would really put paid to cheating online. But it would add so much work to the servers - and probably do some damage to your ping - that the trade-off would make it worthless. Your computer processes that stuff because, well, that's what it's there for - that's why you'll be able to run higher resolutions or put more effects on when you upgrade.