Sony's Hidden Online Strategy

[London Geezer]

The way I see it, most people have 256 Kb upload, right? That's 32 KB a second upload. At 60 fps that affords you 5 KB per frame of upload data. For a client node uploading the gamer's data, that's workable. But if that client node is also passing on data from dozens of other clients, it'll choke.

Until the BW limits are broken there's no amount of clever techniques that can be used to decrease latency, distribute processing in realtime responses, etc. because there's barely enough BW available for a single upstream or downstream for a complex game, let alone excess which you can use to serve other nodes.

5KB x 60fps = 300KB = 2400Kb upload... not gonna happen...

32KB / 60fps = 533Byte

 

[Shifty Geezer]

5KB x 60fps = 300KB

Approximate figures of course. Actually it's appoximately 4 KB a frame. Curses, I can't think of a cunning way to cover up my mistake! Maybe I should just edit it and make like LB's talking nonsense...

 

[London Geezer]

Approximate figures of course. Actually it's appoximately 4 KB a frame. Curses, I can't think of a cunning way to cover up my mistake! Maybe I should just edit it and make like LB's talking nonsense...

Late... I even bolded it


Who says "Curses" anymore!!

 

[patsu]

The way I see it, most people have 256 Kb upload, right? That's 32 KB a second upload. At 60 fps that affords you 533 bytes per frame of upload data. For a client node uploading the gamer's data, that's workable. But if that client node is also passing on data from dozens of other clients, it'll choke.

Until the BW limits are broken there's no amount of clever techniques that can be used to decrease latency, distribute processing in realtime responses, etc. because there's barely enough BW available for a single upstream or downstream for a complex game, let alone excess which you can use to serve other nodes.

I re-read the thread and realized where my mistake was. I missed the MMORPG context totally, and was still thinking about the general case of applying the Cell network concept.

Based on your example and data points, it would be impossible to stuff an elephant down a straw. In a real game implementation, what kind of optimizations are used commonly to maximize the "possibility" ?

* Are data really sync'ed at 60fps rate constantly ? Note that one can also send data out and take a global snapshot independently. e.g., Prioritized message/update may go out to specific members directly for immediate action to take effect.

* In the absence of timely/regular snapshot, how often does game speculate/predict player movement based on last packet received (and then make corrections later based on availability of real data). This may cause players to see slightly different things on screen but the "larger" results remains the same. The global snapshot can contain only "delta" messages to save bandwidth.

* While exchanging data, is it common for nodes to summarize/transform the data so that lesser data gets transmitted (e.g., summarizing individual raw movement data into group configuration pattern for X players and let the local nodes figure out how to show group movements). So for different zones/areas/distance of players, you only need summary info about remote groups to approximate location of players.

As for the larger online strategy discussion, it would not be the first time Sony tried to implement an "Agent Network". The failed Magic Cap system from General Magic, Motorola and Sony would be the first shot: http://www.blakespot.com/nino/html/pdas_ml.html

EDIT: To clarify the 3 bullet points above.

 

[Inane_Dork]

One where Sony only allows sharing of certain types of media. I think I also said twixt my PSP and PS3, not necessarily my PS3 and stranger's PS3.

It would not be an open one wherein I can share any filetype I like., with whomever I like.

Games taking advantage of the "secure P2P" offered by PS3's OS (or whatever) could use it to let me share pre-defined game content with other strangers from within a game, perhaps, but that's different.

That's just my speculation, though, based on what Sony have talked about. The mere mention of "secure P2P" on a slide tells us nothing about how it'll be applied.

I would call that a file sharing service in a heartbeat, but apparently you wouldn't.

 

[Titanio]

I would call that a file sharing service in a heartbeat, but apparently you wouldn't.

In the tradition of open filesharing networks of strangers, in PC-land? I don't think so. As I proposed it, it's filesharing, but it's between me and me (PS3 and PSP), and the filetypes would be limited. Sharing between me and strangers would be rather constrained, to fixed game content within a game. While pretty much anything can be exploited for illegal activity, which I take is your concern, such a setup would limit the possibilities considerably.

That's what I'd imagine would be the most likely scenario, anyhow.

 

[Cattle-Dog]

Who says they spent all that much?

It's not a cheap endeavor:

Mr. Pachter, of Wedbush Morgan, estimates the company has spent more than $1 billion total on the service, including marketing costs. Microsoft won't disclose the cost, though Mr. Greenberg said it has made "a huge investment" in Xbox Live.

>source<

 

[patsu]

It would be interesting to see their breakdown. According to the source article, analysts estimated US$200 mil for XBL development alone, > USD $1b on XBL since its inception, and USD2b for XBL + XB360 over 5 years ?

The XBL number seems too high compared to the XB360 part. Let me know where I go wrong. I would guess the major cost would be bandwidth + hosting, man power and marketing.

Hardware and Licences
===============
Hardware + software costs are usually amortized over 2 years (or more), and recent advance has lowered this ticket item substantially (still tens of millions *list price*). MS should use their own software, and probably practise internal charging here. So part of the $$ spent here goes back to MS.

Bandwidth + Hosting
=============
For bandwidth + hosting of game community services, I don't think it's beyond traditional community services cost. The tight integration between different XBL components may place a tighter load on the servers though (but certainly not hundreds of millions more).

For content download, XBL bandwidth needs may be much higher than Internet file sites due to significantly larger files (> 1 Gb ?). I wonder if they are using bit-torrent in the latest version to alleviate the load.

In any case, we don't know if XBL is hosted on the same infrastructure as MSN. If so, they could leverage on the economy of scale, and even practise internal charging again. If so, part of the reported $$ goes back to MS.

Development & Ops Man power
=====================
All expensive American developers ?

On the community front, development of many gaming support and community functions.

On the gaming front, there should be lot's of effort to integrate, test and support client-server and peer-to-peer network code into many games. This is the tough part given so many implementation approaches.

For ops, there's also manpower to handle billing, system and network admins.

Sales & Marketing
============
Then sales & marketing people to get publishers to support XBL and XBLA. Also to pull relationships to release XBox loyalty card. They may need to inject some money into the loyalty scheme so that the points can be used to exchange for something, or more likely pay an existing loyalty card company to run it.

Plus any Market Development Funds to incent developers and also promote XBL. The "Ring of Light" and other unsuccessful (but probably expensive) campaign $ should come from elsewhere.

Not to forget funds dedicated to XBL primary or secondary market research.

Logistics
======
Logistics for XBL (cards) should be pretty straightforward and can be tapped onto existing Xbox channels.

Customer support
============
Customer support is a wildcard (Usually Passport login issues, billings, and cheat complains. XB hardware support calls should be counted separately), but part of this can be (has been) outsourced based on some negative forum postings.

Others
=====
Mistakes, made by Microsoft that requires rework. I think in the early days there may be rework to fit Passport into XBL. Some large developers such as EA was not comfortable with Passport having access to their customer information (Remember XBL started at ground zero and MS published competing sports games at that time). So Microsoft may have to revamp Passport to handle additional business needs.

It also depends on how they do their accounting.


My gut feel is it is possible to develop and run XBL below USD 500 million if they spend their dollar well, especially now that it is relatively well established. They may be able to streamline it even better.