Do you think SEGA will return in 2010?

Then you have the cost of shipping, the cost of compensating retail for returns, the cost of storage, and most of all the cost of development which must be amortized over an initially indetermine number of unit sales. And of course, you also have the cost of marketing the thing. That alone is easily on the magnitude of 100 million dollars just for the US/Can markets. I also doubt 50 dollars would cover all of the other components in the system, particularly including a high quality LCD that would be competitive to the PSP's screen.

And on top of all that, what media options do we have? Masked ROM is out of the question given its low capacity/cost ratio (9 USD for 64MB from Matrix Semiconductor is the last I've heard on this front) and anything NV RAM based carries the stigmata of piracy and homebrew development that's been known to scare the shit out of publishers everytime it's mentioned as a storage option.

I love Sega, DC and PVR as much as the next guy... but it's over for good. It's all Sony's and MS's game now.
 
I also doubt 50 dollars would cover all of the other components in the system, particularly including a high quality LCD that would be competitive to the PSP's screen.
Click to expand...

What other parts other than the LCD, speaker, controller is there? :LOL:

Like I said the LCD would be a 3.7" QVGA screen. That's not very expensive to begin with, it's rather cheap actually, then when you factor in the hundreds of thousands of units, the price becomes even cheaper. The PSP LCD is expensive because of it's size and resolution, DCP doesn't need a sreen of the same quality, especially when it will be sold for around $100.

Regardless we're just talking about hardware costs and what a DCP would cost to manufacture. Besides your emphasis on marketing budget wil depend on your sales goals. Neo Geo didn't have a huge marketing budget not to mention that a huge marketing budget doesn't guarantee huge sales either. ;)
 
In the end, it comes back to what you said though, Sega/Sammy could sell it for $99 but none of us have an idea what they would lose on it per unit.
 
Sell it for $99 for a loss of $10-$20 per unit initially, that would probably only last for 3-6 months though.

Then the system would start profiting per unit.
 
TEXAN said:
Sell it for $99 for a loss of $10-$20 per unit initially, that would probably only last for 3-6 months though.

Then the system would start profiting per unit.
Click to expand...

But you don't have any idea of what it would cost to make though, so how can you even say it would only be < $120?

That's like me saying, 'oh the PSP only costs Sony $50 bucks cuz they made everything themselves and got a HUGE break on LCDs'.

Sega is my favorite developer but I'm afraid you are seriously divorced from objectivity.
 
I'd been thinking that the computational demands of Sega Sammy's new high-end arcade system would be met by an assortment of processors/processing elements including a hefty math co-processor for mainly physics, and I had thought that Renesas and PowerVR/ImgTec might be the supplier of the system's components outside of the already-announced PowerVR graphics part. After the AGEIA PPU announcement, though, there seems to be a possibility that Naomi 3 is a customer of the new physics chip.
 
Lazy8s said:
I'd been thinking that the computational demands of Sega Sammy's new high-end arcade system would be met by an assortment of processors/processing elements including a hefty math co-processor for mainly physics, and I had thought that Renesas and PowerVR/ImgTec might be the supplier of the system's components outside of the already-announced PowerVR graphics part. After the AGEIA PPU announcement, though, there seems to be a possibility that Naomi 3 is a customer of the new physics chip.
Click to expand...

Yeah thats the exact same thing that went through my mind when I heard of the PPU.
 
The most expensive part of making a hardware platform is the recurrent costs of marketing (which can include the loss-leader business plan of selling hardware below cost), not its development and manufacturing. An arcade system doesn’t need any of that marketing. R&D expense for chip design and fabrication is built into the licensing of those services from dedicated operations which sustain themselves and make it affordable by spreading the cost across an open market of customers. While the high volume of the consumer market reduces the significance of much of the cost for developing a custom processor and thereby achieves a price/performance optimum which leaves R&D of custom processors for the arcade and other low volume markets impractical, custom designs at the system level of arcade architectures can still be optimal for its high performance demands; the boards don’t have to piggyback off the system design of a console.

For their next arcade platform, SEGA probably realized that a high-end system would be better served by heavier specialization of CPU functionality as the demands of game logic, physics, AI, and partial graphics assistance were rising too steeply to make it practical for just one processor. They were probably looking into dedicated math co-processors when AGEIA came along promoting their PhysX chip. It would’ve been a good match, so they may very well have adapted a PPU for Naomi 3. As a result, central processing duties would become more basic, and the housekeeping would be best served by a modest, efficient chip from a roadmap like SuperH.

The reality of the situation may be a while coming, though, with the PowerVR graphics chips set for production later this year and an estimated six or so month wait beyond that -- based upon the schedule Naomi 2 followed -- until the games would be set to go for a JAMMA system unveiling.

The days when console conversions of arcade games were pale imitations may return if Naomi 3 does include such co-processing power and a likely 1+ GB RAM, and provided the developers don't hold back for the sake of easy porting.
 
The four most computational tasks for a 3D game are:

- AI
- Physics
- Transform and Lighting
- Rendering

The ultimate system is the one which has a dedicated processor for each of the above tasks.

Imagine if Naomi 3 had the following

AI - SH-7?
Physics - PhysX chip?
Transform and Lighting - ELAN 2?
Rendering - PowerVR Series-5? x2? x4?

Then imagine if each of these units had dedicated ram say for example 256 or 512 GDDR4 each.

Then you are looking at a trully monstrous board.

It would basically be the architecture of the Naomi 2 expanded upon.
 
If Sega were going to produce a new Arcade board, why not base it on high performance 'available' parts...

dual Athlon or P4 motherboard,
dual Geforce6800( sli ) or equiv.
Physics coprocessor board ( as shown at GDC )

They could install a custom microkernal ( linux or something ) to get 'bare metal' performance, or just use a stripped down WinXP kernal.

Something fully custom obly makes more sense if they are making a console..
 
Crazyace said:
If Sega were going to produce a new Arcade board, why not base it on high performance 'available' parts...

dual Athlon or P4 motherboard,
dual Geforce6800( sli ) or equiv.
Physics coprocessor board ( as shown at GDC )

They could install a custom microkernal ( linux or something ) to get 'bare metal' performance, or just use a stripped down WinXP kernal.

Something fully custom obly makes more sense if they are making a console..
Click to expand...

Why? Those arcade boards will need to run a handful of games (namely, a racing game, Virtua Fighter 5, the next House of the Dead, and some other titles). It makes sense to manufacture something that Sega will know can run those games in the best possible way, at the price they wish to pay.
 
Crazyace said:
dual Athlon or P4 motherboard,
dual Geforce6800( sli ) or equiv.
Physics coprocessor board ( as shown at GDC )

They could install a custom microkernal ( linux or something ) to get 'bare metal' performance, or just use a stripped down WinXP kernal.
Click to expand...

If thats how they were thinking then they would have created chihiro 2 instead by licensing X360 from MS. Probably would have doubled the ram like the original.
 
A Sega Sammy arcade system needs the price/performance of an embedded architecture to be able to justify itself by significantly outperforming the embedded designs of consoles, and there'd still be some kind of expense associated with the licensing of someone else's console design.
 
For a high end arcade cab, it doesn't have to match the price / performance of a console - it can use more expensive parts ( for higher quality ) - and if Sega leverage high end PC parts they skip a lot of R&D expense.

2 high end (512MB maybe) SLI NV chips ( or X850 ) will probally match the XB2 graphics, and a dual processor top end Intel or Athlon should be more than a match for a console cpu - as long as that is your only target. After all at the end of the day a console is built for a very low pricepoint, and an arcade board could easily be 5-10x the cost.
 
That example actually shows that an expensive PC-like system still doesn't meet the critical requirement of substantially outperforming the consoles yet.
 
You must log in or register to reply here.

Similar threads

W
E3 2010 - Here Are My Plans & What Else Do You Think I Should See?
2
Replies
27
Views
4K
AzBat
A
M
Nintendo Year End Financial Briefing - Iwata Q&A
Replies
0
Views
2K
Megadrive1988
M
M
rumblings of something huge going on with SEGA
11 12 13
Replies
252
Views
26K
Zapp$ter
Z
M
Rumurs on Sega's *possible* Hardware Re-entry
2
Replies
35
Views
4K
Deepak
D
A
NVIDIA Analyst Day: Webcast Report (+PSP2 Hints)
2
Replies
36
Views
9K
Geo
G
Back
Top