Gaikai co-founder David Perry reveals his ambition to take Gaikai to consoles
David Perry has revealed he is eager to see the cloud-computing service Gaikai available on consoles, and thinks Nintendo would benefit from the service in particular.
In a Develop Blog Post – which looks at the genesis of the Gaikai model – the Perry explained,
“What I would love to do is have the hardware manufacturers support our embedded streams. I’d like them to support our custom stream and then allow live demos of the games on the consoles themselves.”
Perry explained that the Gaikai service – which essentially recieves player-input across the net and streams video footage of gameplay – could be embedded into the platform holders’ digital distribution stores.
“Just imagine being on the PlayStation Store, looking at all the games and trials, and being able to play a demo of the game, instantly, without downloading,” he said.
Advertisement
Perry added that Gaikai was a service that publishers and platform holders could take control of. A demo of a game could last as little as 48 hours, or up to a certain point in the game, before users had to download.
“Or,” he added, “they could charge you to continue; 100 Microsoft Points to carry on through the game.”
Though Perry foresees Gaikai dramatically changing the console demo process, he implied that the online-cautious Nintendo would best benefit from the system.
“I actually think Nintendo could do real damage with Gaikai,” he said. “They’re now paying attention to the draw of online gaming, and they also have some of the best games to get into quickly.”
Perry first told Develop in a recent interview of the key difference between Gaikai and cloud computing service OnLive, inasmuch as the OnLive service is positioned as a rival platform to the platform holders.
Perry’s Blog Post emphasises the game design veteran’s belief that Gaikai can do the opposite; assist the platform holders in reaching out to new audiences.
... and this was what I found.
From the Gaikai website.
"As YouTube and a multitude of copycats have shown, streaming videos online in a web browser is fast, efficient and reliable"
Except it isn't, streaming video is reliable if you have a fast and consistent stream. Try watching a Youtube video in high quality with a 512Kb/s connection and you'll get something far from "fast, efficient and reliable". If everything was so "fast, efficient and reliable", we'd never see a video buffering before/during playback. I see this with a 2MB/s connection on Youtube. Are these people trying to tell us that they can stream lossless HD (1920x1080) gaming, without lag or buffering on less than a 2Mb/s connection? Yeah, sure.
"in a few milliseconds. Your input – every keystroke, every mouse click – is sent back over an encrypted channel to the game"
A few milliseconds there, a few milliseconds back, we're suddenly not talking about a few milliseconds anymore. Try using a VST with even a 50ms delay between input and sound and you'll notice how detatched you suddenly feel. Now imagine that situation with input and visual feedback as well. Anything more than instant is going to be unacceptable, and the fact is, current networking can't provide "instant".
Then let's do some maths. Imagine I'm sat here on my 2Mb/s connection, with a 10GB per month bandwidth allocation (which isn't particularly rare in the UK). I want to play the latest game for 8 hours or so, so I start playing.
Now 2Mbs equates to say 220KB/s, or 1MB per 4 seconds, or 15MB per minute... or more critically, just short of 1GB per hour. So in my 8 hour session, I've used up about 90% of my monthly bandwidth allocation.
So when you read their statement "So there’s your choice: a costly state of the art DVD game trapped on your machine – or a dated browser game you can play anywhere." They neglected to add the option of "a game you can play one day out of every month".
Even if your allocation isn't 10GB a month, I dare anyone to tell their ISP that they're going to be streaming a constant 1GB per hour, for 4 hours a day, every day of the week. See how long it is before you get dumped on the low-rate connection for abusing the "fair use" policy. You really think they're going to let you stream that much data, at the expense of other users, on a regular basis?
All these online services, Onlive, Gaikai etc... are speculative nonsense besed on presumptious scenarios that don't exist. We don't have universal fast broadband. We certainly don't have unlimited bandwidth to go with it. You might as well just advertise, global online, 1 million player, lag free gaming, because it's theoretically about as credible as Gaikai's claims, just not on this planet.
I don't know about clouds, but this sounds more like "pie" in the sky every time I hear more about it.
On another well-known site that we can't link to, they have an interview with David Perry about how Gaikai is great for investors.
Anyone else seeing a written off Ferrari?
Among companies like Gaikai, Onlive, Otoy, Playcast, and probably other big names eventually (like Verizon, Apple, Microsoft) there will be plenty of momentum for solving the technical issues involved in streaming video game experiences.
What is more interesting at this point isn't the technical, it's the commercial. Which of these companies can make a commercially viable effort? Gaikai's approach is interesting - if Nintendo offered a way to test-drive the games by streaming a demo, it would be interesting to see if that drives up sales for those games. I could see something like that as a service the Nintendos of the world offer to their third-party developers to help boost their sales.
The problem for Nintendo in this scenario is the controllers (wiimotes) themselves probably introduce too much lag - but that's a technical issue I am convinced is solvable. The question is who is going to accomplish this first, and from what commercial approach?
