Sony in talks on commercial use for PS3
- Thread starter patsu
- Start date
I told you so
Why SPE has a small local memory is not only because of the memory hierarchy, but also because it's suitable for distributed computing. The old APUlet is still alive as SPUlet in this patent application in 2005
http://v3.espacenet.com/textdoc?DB=EPODOC&IDX=US2007074206&F=0
The security sandbox of SPE is useful not only for DRM, but also for tamper protection in distributing work units onto random nodes.
The F@H client for PS3 was I think basically a demonstration project to show PS3 consoles are a manageable computing resource that are ready for commercial use, it's not that they suddenly became philanthropic. Also the pie grows every month. For a cool GUI or a mini game PS3 users are likely to return their watts. SCE can sell a place in XMB or be paid to build a GUI client distributed for free. So in this new economy it's like buying entertainment not by actual money, but by watts. It's a form of utility computing about which Kutaragi was talking in an interview or a speech. Currently it's just a simple form such as F@H, but in future, more complex form of this system will appear on PS3 like mixing distributed computing workload into an actual game or forming a P2P cluster.
Why SPE has a small local memory is not only because of the memory hierarchy, but also because it's suitable for distributed computing. The old APUlet is still alive as SPUlet in this patent application in 2005
http://v3.espacenet.com/textdoc?DB=EPODOC&IDX=US2007074206&F=0
The security sandbox of SPE is useful not only for DRM, but also for tamper protection in distributing work units onto random nodes.
It won't work without either being a background operation that people don't mind running, or having some incentives (could be paid in Sony network credit). Sony's petaflop dreams have been dashed by a lack of interest - 80,000 PS3s, a quarter on at a time. There's lots of processing power there, but that's from people with an interest in the Folding cause. Who's going to want to donate Cell cycles to some corporate startup or pharmaceutical compnay (and we all know what great reputation those compnaies have!) making money from their research?
Me and by the looks of it i´m not alone, if the Folding Client was a truely background working program that didn´t intrude in anyway more would run it as well.
Do you think your negative tone reflects the FOH guys as well, they doubled their efforts with the PS3 and it can only go up.
I don't think you read my post properly. Folding is a 'charity' good cause. It's open knowledge for everyone. People like that. Would you rather instead of running Folding or other good cause clients, run commercial ventures for companies?
I think it's suggested that you would be compensated for your contribution in some way. That's the incentive. You'd then have the question of whether people would be charitable with their CPU cycles and stick with things like F@H, or be more tempted by commercial reward.
The incentive is how the client is built. 1 SPE is reserved so it's enough for donation, but 2 or 3 SPEs won't hurt. With the left SPEs, you can build a game PS3 users can play for free. While users play that game online, computing resources are donated.
It's similar to in-game advertisements in the way that it donates a part of computing resources in a game console for a non-gaming purpose.
So in a practical scenario
pro
The cost to build/lease a good GUI client (which is a visualizer, a game, or whatever that is entertaining) can be much cheaper than buying a hardware computing cluster. A hardware computing cluster gets obsolete quickly, but the PS3 install base grows every year.
con
It's unpredictable how much computing resources can be obtained. No one may play it if other more interesting things appear. (But this can be amendable by not associating a specific client with a specific workload)
Last edited by a moderator:
Of course, but there can be a case where collective payments get larger than the cost to buy an actual hardware supercomputer.
Yeah in that case paying each contributer is even more difficult. Giving them rewards in the form of software (game, movie, music, in-game item) is more affordable and competitive. Anyway supercomputers with Cell or Cell-like processors are still not so common.
It'll be like Google's advertising model where the highest bidder gets the largest resources.
archangelmorph
Veteran
I think there's alot of potential here..
Sony could do so much with it in terms of commercial potential to business and incentives for users..
Some ideas:-
A) Simply setup aditional schemes where users can sign up and offer idle time for commercial distributed computing.. The users could benefit from earning incentive points which could be redeemed for real world vouchers for specific products, points for Son'ys online store and more.. Problem is unless a global platform was established, seperate clients for seperate companies could swamp the system..
B) Same as A) but where the client offers some kind of game platform where users can download and play games whilst the distributed processing goes on in the background aswell as full Cell use during idle time.. (thus encouraging users to spent more time overall with their PS3s in use..)
C) Setup a platform the same as B) but tie it into the whole Game 3.0 idea by allowing user-created games to be put together and deployed onto the platform which allows users to sign up and play homebrew games built to a restricted Sony-defined spec leaving the rest of the system open for background processing throughout..
The best option would be to take the elements of all three and combine them into another great service (along side Home) which the PS3 could offer to add value to the consumer and to Sony themselves..
Sony could do so much with it in terms of commercial potential to business and incentives for users..
Some ideas:-
A) Simply setup aditional schemes where users can sign up and offer idle time for commercial distributed computing.. The users could benefit from earning incentive points which could be redeemed for real world vouchers for specific products, points for Son'ys online store and more.. Problem is unless a global platform was established, seperate clients for seperate companies could swamp the system..
B) Same as A) but where the client offers some kind of game platform where users can download and play games whilst the distributed processing goes on in the background aswell as full Cell use during idle time.. (thus encouraging users to spent more time overall with their PS3s in use..)
C) Setup a platform the same as B) but tie it into the whole Game 3.0 idea by allowing user-created games to be put together and deployed onto the platform which allows users to sign up and play homebrew games built to a restricted Sony-defined spec leaving the rest of the system open for background processing throughout..
The best option would be to take the elements of all three and combine them into another great service (along side Home) which the PS3 could offer to add value to the consumer and to Sony themselves..
But that's the thing, isn't it? A competitive pricing point will be found, or not. Whether this is through combining it with game-sales, Sony wallet rewards, or whatever else, doesn't really matter.
Say that it costs me 14.4 euro to keep my PS3 on for a month. Someone is willing to pay me 30 euro to keep it on when I am not gaming (for, say, 12 hours per day) and have it contribute towards their project, that could be worth it for me.
Of course, there are areas in the world where electricity is cheaper.
That would mean 60 euro for having a PS3 available 24/7 for a month (2x12 hours). 1000 PS3s would therefore cost 60000 per month. I'm taking fairly generous numbers here, I think.
How would that compare to hiring a super-computer with comparable performance?
Even if I got only 14.4 euro per month, that would still cover my electricity bill and still give me 12 hours per day to use my PS3. So I might still do that, in which case monthly cost would be 30000 for 1000 PS3s. Etc.
It all depends on whether or not there 1) is an alternative and 2) is more affordable in terms of real costs (programming, operating, logistics, etc.)
Heh heh... it doesn't have to be games though. PS3 has plenty of free cycles while playing music and DVD too. Don't know how well it will turn out.
It looks like Sony is consolidating/repackaging Internet into the Playstation brand.
* Distributed/Grid computing (F@H and now this)
* MySpace, SecondLife, ... (Home)
* P2P media/file sharing (Downloadable game sharing, Playstation Home video sharing)
* YouTube, Grouper (Home)
* Web surfing
* RSS
* Remote access (Location Free, Remote Play)
* Online gaming + built-in specialized community (LBP, SingStar, and other Game 3.0 related effort)
* Media sales like iTunes (later this year)
We may see PayPal/eBay (Consumer-to-consumer commerce) in the future too.
It looks like Sony is consolidating/repackaging Internet into the Playstation brand.
* Distributed/Grid computing (F@H and now this)
* MySpace, SecondLife, ... (Home)
* P2P media/file sharing (Downloadable game sharing, Playstation Home video sharing)
* YouTube, Grouper (Home)
* Web surfing
* RSS
* Remote access (Location Free, Remote Play)
* Online gaming + built-in specialized community (LBP, SingStar, and other Game 3.0 related effort)
* Media sales like iTunes (later this year)
We may see PayPal/eBay (Consumer-to-consumer commerce) in the future too.
While interesting, distributed or utility computing schemes have been thought of before.
I think Sun's tried pushing something like that to little effect.
I don't know if the PS3 can do well for a commercial scheme if the entire platform didn't already have that as a primary goal.
I think Sun's tried pushing something like that to little effect.
I don't know if the PS3 can do well for a commercial scheme if the entire platform didn't already have that as a primary goal.