That would be my thought as well, that the Arcade remains a loss leader. It would seem a little strange for the SKU not to be a loss leader and have the Pro cover the spread. Beyond that at the end of the day I don't think that cheaper though it be, the Arcade at this point would actually have ~$100+ less worth of BOM than the PS3, where relative to the higher-end SKU it's probably all of a ~$40 BOM differential.
NPD December 2009
- Thread starter RobertR1
- Start date
That would be my thought as well, that the Arcade remains a loss leader. It would seem a little strange for the SKU not to be a loss leader and have the Pro cover the spread. Beyond that at the end of the day I don't think that cheaper though it be, the Arcade at this point would actually have ~$100+ less worth of BOM than the PS3, where relative to the higher-end SKU it's probably all of a ~$40 BOM differential.
Last year it was stated by Alan Greenburg (I believe) that they broke even on the Arcade 'with accessories' which I take it means extra controllers, HDDs but not Live points cards as I don't see a need for someone with 512MB of space to have those. I interpret the comments to mean 'The Xbox 360 Arcade costs us $220-$240 to produce'.
The problem with trying to guesstimate the Xbox 360 Arcades BOM is that we haven't seen a recent breakdown of costs even one from iSupply.
Its been a year since they made that statement. The yields on the 65nm process must be pretty good, they'll probably be throwing out fewer packaged chips due to yield and the cost of the 65nm wafers must be pretty good too now. The bill of materials for everything else must also be lower due to reduced fuel prices and lower commodity prices for raw materials like aluminium and copper for the heatsink and other PCB materials. Lastly it depends on how the royalty rates scale with time, the chips have been in production for 5 years so the I.P. behind them must be getting cheaper too.
Obviously the loss-leader estimate is sans accessories; with accessories everything is a different ballgame. Anyway certainly the yields on 65nm must be very decent now, but they were never so horrible to begin with when we are talking TSMC's 65nm process. More important will be how/when they transition to either 55nm, 45nm, or 40nm. Or did they already go to 55nm on the GPU? So many threads stuck in my head...
As for the ancillary/back-end costs, I would not think too much with regard to copper prices, oil costs, etc... because they are all of them tied in part to the fact that the US dollar was weakening as well, and if the currency weakens for MS, then their outsourced costs respectively go up. I would call that area a wash and simply say that there are too many factors on the back-end for us to know with certainly which direction it ultimately nudged costs in. I love talking about just those kinds of factors, so no unwillingness - simply advocating for brevity.
IP is per part - it should not cost reduce I do not believe... at least that is what I would assume on the contracts written. And so really I think the gains to be had are almost purely silicon and market forces related.
For myself I am going to continue assuming that the Arcade is sold at a loss. To have two SKUs both profitable would seem a little strange, since you could obviously lower one into loss-leader territory in order to skim a greater buyer pool at the same time that you hedge against the costs by still being hardware profitable in the aggregate.
Yeah, accessories are a very interesting topic. Sony gave up many millions of profit when they dropped the obligatory memory card of the PS2 from the new system and included a hard drive by default
In this way the Arcade might be a very good business - I imagine if 25-50% of the users decides get a hard drive eventually, it's already enough to cover MS's losses. I guess we don't have any data on the HDD sales...
In this way the Arcade might be a very good business - I imagine if 25-50% of the users decides get a hard drive eventually, it's already enough to cover MS's losses. I guess we don't have any data on the HDD sales...
Even with say a difference of 10% in yields it does make a reasonable difference in cost. They don't have more than one SKU of chips in this metric, they are either good or bad so therefore they probably cannot reach the same level of yield per wafer as a GPU manufacturer who can bin for different markets. The chips are either suitable or not. They pay for packaging and testing whether the chip works or not, so the cost is not just on a per die basis rather a per package basis. With the GPU the package includes both an ED-Ram and a GPU chip so yields are even more important for this relatively more expensive package, which is why the GPUs of the PS3 and Xbox 360 are typically a year behind the CPU in progressing to a new node.
The Xbox 360 is using Chartered Semi conductors 65nm process for the CPU, NEC 80nm ED-Ram and TSMC 65nm for the GPU IIRC.
As for the ancillary/back-end costs, I would not think too much with regard to copper prices, oil costs, etc... because they are all of them tied in part to the fact that the US dollar was weakening as well, and if the currency weakens for MS, then their outsourced costs respectively go up. I would call that area a wash and simply say that there are too many factors on the back-end for us to know with certainly which direction it ultimately nudged costs in. I love talking about just those kinds of factors, so no unwillingness - simply advocating for brevity.
You're right here I guess, its a complicated issue relative to its worth in discussing it.
I would assume the opposite really. The consoles are designed to be expensive to start with but come down in price as the technology stales. More advanced technology tends to be more expensive to license and cheaper technology tends to be cheaper. For example the cost of BRD playback on the PS3 has come down to $9.50 if they pay market rates for it.
The cost of the Xbox 360 is also dictated by their intentions with their Natal interface. They cannot easily reduce the price and then increase it once they bundle the controller with the system. In addition to this one can assume the Arcade is the SKU which also earns them the lowest revenue per unit sold. Ideally they would want to break even on this SKU as it has the lowest earnings potential of their two SKUs.
...
True but the Xenos itself does have a redundant shader array, so it can withstand die defects to a certain extent by design.
The former on the packaging yes, but the PS3 GPU isn't a node behind because of any other reason than it uses a completely different node to begin with, namely Toshiba's CMOS. When the OTSS fabs get far enough along to switch, they do. No driver determines when RSX changes nodes other than this, and I believe we should be seeing 45nm soon on RSX.
There's a thread around here where we were analyzing the Xenos on a half-node shrink; maybe that was when it was hanging out at 80nm, who knows (though probably then also). I tried searching for the thread, but couldn't find it. Maybe someone else has it on tap, though. Anyway if I am confused, then it is in thinking that discussion was in relation to 65nm-->55nm vs 90nm-->80nm. Whichever it was, I was pretty on top of it at the time, I remember that!
The consoles, yes, but the agreements struck with the IP firms at the outset are generally fixed-cost agreements. The console can get cheaper afterall without the IP costs for the processors going down - that is why they are licensed to begin with, so that the console makers can control the bulk of the costs on the fabbing side.
Moments in time though. I'll admit to not having any idea as to what BD's playback licensing costs are beyond the hardware itself at the moment - if it is $9.50, well - then it's $9.50. Obviously when MPEG LA and the BDA were hashing this out post HD DVD defeat, it may have settled at a lower spot than initially when it was a bunch of competing claims of IP property; with the BDA it is an aggregation pool vs a one-on-one contract with a party like NVidia or ATI. The licensing costs of BD may come down over tiem simply to allow for a wider adoption rate of the technology as they seek to lower the price of entry along the demand curve; for NVidia and ATI though, I just do not believe there would be any sense in allowing any sort of IP floatdown down the line.
There are members around here that definitively know the answers to this question of course; we can only hope for their interjection.
Well and either way you can't just drop the price of your SKU every couple of months - it would just seem too fast. But with Natal I don't know that I believe it will be an automatic-inclusion with every Arcade system... I could still see a standalone Arcade, and then the Natal bundles on top of that. $199/$179, $179/$149, $199/$149... I would believe anything.
My take on Natal by the way is more akin to Guitar Hero than it is to a re-envisioning of the console space. I see it as an accessory that in its own right will sell millions and be profitable, and it will play into a specific genre of gaming.
This is true to some extent. The Xenos chip has 51 shaders and 3 are redundant. However the modern ATI architectures have both fine grained redundancy as well as the ability to turn off whole blocks including texture units. If a fault is situated in say a texture unit it doesn't help them that much.
Ahh I didn't know this, I figured it was TSMC.
There's a thread around here where we were analyzing the Xenos on a half-node shrink; maybe that was when it was hanging out at 80nm, who knows (though probably then also). I tried searching for the thread, but couldn't find it. Maybe someone else has it on tap, though. Anyway if I am confused, then it is in thinking that discussion was in relation to 65nm-->55nm vs 90nm-->80nm. Whichever it was, I was pretty on top of it at the time, I remember that!
There was some speculation that Microsoft shrunk the Xenos chip on the 80nm process to help deal with the RROD issues early in the life-cycle of the console. This was indicated by the reduction in die size for their Falcon revision.
Well I'll defer to your greater knowledge in this area. Maybe Dave will tell me? Haha, that'll be the day!
Thats true, different markets and different expecations.
There are members around here that definitively know the answers to this question of course; we can only hope for their interjection.
Dave knows, maybe if we beg him simultainiously he can lay open the trunk of his chest of secrets?
My expectation is that Natal is a 'relaunch' of the system. Since its coming around the time we would expect their 'Valhalla' 45nm combined chip to be released it could coincide with a slimline console as well. They couldn't fit the camera into the box otherwise and its easier to deal with a smaller box than trying to stuff a larger box into their channels. They could probably afford to bundle the camera with a slimline console if it follows along the same lines as the PS3 slim price reductions.
The noise surrounding the interface and the sheer number of rumoured games coming out around the time of Natal which are to support the interface suggest otherwise. If nothing else Natal would be a killer multimedia interface.
It's probably the 80nm shrik I am thinking of - especially since I can't be thinking of a shrink that hasn't happened yet.
That thread's around here somewhere in any event....Well I would sort of expect a slimline version of the 360 as well, in some form or another. It's sort of a weird shape right now, so it will be interesting to see what form it takes. As for Natal, I do agree it would be a great multimedia interface; what surprises me of course is that the PS3 has both had the basics of this interface since inception and has hardware ideally suited to the processing required, yet the EyeToy utilization and expansion has languished for years. Bizarre. Beyond the 'guitar hero' type success I'm envisioning, maybe MS really will popularize it as a non-gaming interface. At least then I suppose I could hope that Sony actually devote resources back into it's own products!
corduroygt
Banned
I predict 360 arcade + natal to price match whatever the Wii is going for at the time, no more no less.
It's probably the 80nm shrik I am thinking of - especially since I can't be thinking of a shrink that hasn't happened yet.
I banged off a PM to Charlie, so hopefully he can tell me something more about the royalties.
http://news.bbc.co.uk/1/hi/programmes/click_online/7820866.stm which was CES 2009. This shows a possible direction for the Xbox 360 hardware and you should note the NXE interface. They may even be able to integrate the Xbox 360 with TVs which would make a hell of a lot of sense from both the perspective of a TV manufacturer struggling with low margins on their products and from the perspective of Microsoft wanting to expand their market reach. Its all about sharing the Live revenue...
The problem for Sony wasn't that it was too hard to do in hardware but its quite a tricky little thing to do in software. Microsoft has years of experience with interface and dozens of people working directly on this application. They apparantly have algorythims (spelling) which are years ahead of the academic world.
Lastly I think the major advantage Microsoft has with Natal over both the Arc? Gem? and Nintendo's Wiimote is that they acknowledge the flaws of the next generation interfaces which are present at the moment. It has to be cheap enough to augment the experience without expecting to replace it. I would suggest that the Wand may be simply too expensive (for what it does) and the Wiimote cannot do everything.
It'd be worth listing your perceived flaws and how Natal addresses them. Certainly a fair bit of what I've seen about Natal isn't about addressing the flaws of current interfaces but extending interfaces, and in some pretty silly directions sometimes. Arm waving to control a TV, when instead I can press a button? No thanks! The dual-stick controller is very good for the games it controls. SuperStardust and U2 are better games on DualShock then they would be on Natal, I'm sure. Natal needs a whole new branch of design to make it worth having IMO.
I'll be gobsmacked if a big round light and a few motion components costs notably more than Natal. I expect it to cost less (though of course there's the PSEye to worry about as well). And Natal hasn't yet shown it can cover all the bases, so key areas in existing games where the wand will work well, Natal maybe won't, at which point you'll need additional costs (eg. Wielding a virtual gun in game).
If you can think of any flaws in the current Wii Mote with Motion + then you can pretty much apply them to the Sony wand with the addition that the Sony controls require a purchase of at least 4 (multiplayer is a must) wands and a camera on top of the price of the PS3 itself. Considering the nature of the core offerings on the Wii the flaw I see is that if the games haven't really worked on the Nintendo machine with the interface built in, how will they find it on the PS3 without the interface built in?
Because Natal doesn't attempt to address the exact same market as the Sony Arc it can prove its worth to people who aren't sold on the Wii interface. So whilst you may not want to wave your hand at the TV to change the channel, you don't always have a controller within arms reach or have a desire to reach for one if say your hands are dirty, such as preparing food as an example.
Beyond the media interface you can get some completely different game ideas. There are popular board games like Trivial Pursuit, quizes etc which would work great with Natal. Theres educational software for children to aid learning (those grade 1-5 PC games I believe), Adult educational software for people to learn to speak/read/write another language. Im not thinking of core games here but rather improving experiences which can be shared with the whole family in a group setting.
One camera vs camera + 4 dual shock wand controllers?
Suggesting that the cost of the Wand is inextricably linked to having to buy 4 of them is the same sort of fallacious 'price comparison' that Sony pulled on us constantly when comparing value-adds. ('Well, clearly every gamer wants Blu-Ray and Wi-fi, so that puts the price of the 360 way above that of the PS3.')
It's difficult to have great hopes for the Wand; the advantage is that the PS3 already has a large 'predictable' audience, something that has mostly eluded the Wii. So it's possible that gimmicky solutions like RE5's pointer-based controls will see some measure of success. Of course, adoption hinges on the whole 'add-on vs. core device' debate, which doesn't really favor anyone but the Wii, really.
It's difficult to have great hopes for the Wand; the advantage is that the PS3 already has a large 'predictable' audience, something that has mostly eluded the Wii. So it's possible that gimmicky solutions like RE5's pointer-based controls will see some measure of success. Of course, adoption hinges on the whole 'add-on vs. core device' debate, which doesn't really favor anyone but the Wii, really.
I don't think Natal is just a bog standard camera.
Natal's main strength IMHO lies in its ability to work without the hands which can keep holding the traditional controller.
I don't know if there's any research on this field but one of the most obvious looking uses to me is to add leaning into FPS games by tracking the upper body motion. I mean everyone has been leaning to the sides trying to avoid fireballs in Doom, it's such a natural and even unconscious reaction, that should be fairly easy to track. It would increase the immersion a lot, it would be fun... I really hope that at least Reach gives it a try. Then there are probably other options as well, including head tracking, and you also have your feet that could be visible for the camera... Sure, it'd get confusing after a certain point, having to use your entire body and the controller, but it should work at least to some extent.
To put it in one short sentence - Natal can also work as an addition, an extension, to traditional hardcore game controls. This would allow it to reach a larger user base then just casual games, without requiring the developer to support it exclusively.
And the beauty is that neither Sony's nor Nintendo's approach can compete with that.
