Didn't Sega do this back in the 80s or something? Are we sure this is that big of a deal?

I think certain games could probably do this... but yeah I'm struggling trying to imagine games like Gears of War doing this with reasonable latency. But still I'd be interested to see it in action.. I wonder if they will be at E3 this year?

Well.. I thought about this some.. as long as they "delay" the screen response based on your button response.. you can "fool" the client into showing a sync'd response (you'll have a tiny lag..but it might be so tiny you wouldnt even notice it).

Hmm.. this product actually looks really close to something we looked at 2 years ago at Showtime of all places.. I wonder if its the same people and they either got spun out or bought themselves out of Showtime. I'll have to do some digging around.

Unless they throw every possible action down your connection simultaneously (e.g. Go left, go right, jump) and then only show the appropriate one at the client. But imagine the bandwidth then. O.O

Wouldn't they have to then have different versions of their micro usb box thing? I think this is quite interesting, but I do wonder if it's a move to get someone to buy them out or licence them. Just waving a threat round in the air.

Worst lag ever.

I don't get it. Lag and all that notwithstanding, if I wanted to have the equivalent experience that I'm getting right now on my PC with this service, wouldn't there have to be a PC somewhere in serverland with the same guts as my current PC? I get that I would essentially be "renting time" on that PC, but I'm not sure where the enormous amounts of cost savings come in, since they'll have to be running tons of high end PC's/consoles along with the bandwitdh.

Some lag where an enemy isn't where you thought it was, or maybe you didn't quite make that jump across the chasm, these are things people can live with. But even the most casual or accepting gamer won't tolerate their entire experience being lagged. Imagine playing Quake where you move the mouse to the right with mouse look and it takes two or three seconds for the view to pan.

I worked in a test lab for a large company, and we often had pressure to facilitate remote testing. Even though we were not doing anything complex with the applications we were testing, many testers hated doing any work remotely. Even under ideal circumstances, the latency (due to the limits of the speed of light and all the switches getting their hands on the packets) was bad enough to make doing even basic tasks seem really painful. Waiting for a mouse pointer to catch up even under 75ms of latency was painful for them.

This is, of course, due to expectations. People expect things to work a certain way and to get feedback instantly. This is why having a little lag in an online game is tolerable. Because even when it happens, your feedback from what you do, and see instantly on your screen, is all working like you expect.

I don't buy it, except for something like Peggle, or trivia and card games. Unfortunately you simply cannot defeat the laws of physics so unless they stage their servers in every neighborhood across the country there's simply no way around input lag of dozens to hundreds of milliseconds. That might work for things like Killzone 2 -- doh! I went there... -- but for the kinds of games they appear to be targeting which are high-end, rig-crushing FPS games, I just don't see it.

This is nothing but another Phantom.

It will be riddled with every kind of DRM you can think of.

Once publishers have control over distribution, like this machine, they will be able to charge whatever they want - expect $60 DLC games.

This kind of thing is only being promoted, partially, to kill piracy, and largely to kill the used game market.

The video quality will be rubbish.

There will be considerable lag.

There will be issues due to the monthly bandwidth limits that most ISPs have in place.

There will be situations where all of their render machines are filled up, and users will have to wait in line to be able to play a game.

This company will be out of business within three years.

I vote scam until I hear reports on actual use of actual product from people who have no reason to lie about their experience.

Even if their technology is as impressive as claimed, I remain highly skeptical of how efficient it will end up being. ISP's are all differently complicated, not to mention that many of them are instituting new types of bandwidth restrictions all the time. I can imagine a bunch of rough edges to work out. Imagine not being able to play any of your games for an entire weekend because a server room in Rochester is over capacity. Even if it worked flawlessly, I think it won't be a hardcore gamer thing anyways since I guarantee you it won't be able to handle the typical widescreen monitor resolution, at least at launch.

Also, I would definetely wait for a while. From what I understand from the various articles (most of which sound like re-written PR), you pay an annual fee, and then you buy/rent games from the service. It's already an uphill battle convincing the average gamer to spend $50 on a game off something like Steam where you don't get the physical box, but at least then there is the ability to save back-up files and the potential of a crack should Valve dissapear. Imagine how comfortable you would feel spending $50 for a game that only exists off of a server and dissapears once this start-up company that spent 7 years on R&D potentially goes out of business, or even if you stop paying for the service.

I don't have a link handy but PCWorld just did an article on this where they played Crysis at 720p with this thing. It sounds like they are worried about lag as well (you need to live within 1000 miles of one of their data centers). Also the pricing. They are charging for demos as well as full games as well as a subscription fee.

The backend costs to run this thing would be completely insane. At my last job we ran projections for delivering HD level content for teleconferencing and it just didn't scale well at all even for one server farm. Having to have multiple farms, with load balancing & redundancy, delivering potentially hundreds of GB per minute isn't something that will be paid for by a $15 a month subscription or a few $60 games.

Good luck to them, it's a neat idea but unless they have some serious investment back I predict this to be another vaporware announcement.

This will never work, because the cost doesn't scale in any sane way. Lets make some assumptions:

- Every player requires at least 1.5Mbps in continuous traffic.
- You have an impossible 5ms latency to the server.
- You are playing a multiplayer game.
- The server hosts more than one player.

The bandwidth required would absolutely dwarf Hulu and and Youtube. Gaming is not a movie that can be downloaded once and cached - it's a continuous process, so the full video bandwidth is required as long as the players are playing. A 1-Gbps network connection, at full speed, could conceivably support between six hundred and seven hundred simultaneous players at their stated bandwidths, only counting the lowest bandwidth necessary for video. Given 50% utilization of the 1Gbps connection, and assuming 10 cents per gigabyte of bandwidth, that puts your bandwidth costs at:

500 Mbps/8 = 62.5MB/s
62.5MB/s * 86400 sec/day = 5,400,000 MB/day
5,400,000 MB/day / 1024 = 5273GB/day
5273GB * $0.10/GB = $527 per day

So, at ten cents/GB, which is a very good bandwidth price, you are paying approximately 87 cents per day per user just for bandwidth. That's $26/month per user, or $312 per year, per user, JUST FOR BANDWIDTH at the lowest settings and the best possible bandwidth price.

That puts the floor cost of the service at $26 per month, without accounting for any other costs at all. Even if you assume the fantastic rate of 5 cents per gigabyte, that is still $13 per month, per user, of direct overhead cost, without ANY other infrastructure - which of course will be required, because the amount of bandwidth is insanely high and cannot be supported except by the largest providers. ALL of this bandwidth must be very low latency, or the games will be unplayable. Rack space, hardware costs, monitoring and maintenance, salaries, high-end networking equipment, community service - there's no way they can do this and get out of it for less than $25 a month, and more likely $50 per month cost. There's no possible way they can make money off of this.

I don't agree with the "this could never happen", but I certainly agree with the "this ain't it" crowd. It's a conceivable product, but it's not currently financially or (to my knowledge) technically viable right now.