What is the true cost of manufacturing the PLAYSTATION 3?

[MBDF]

Hi, I was wondering If we could find out the true cost of manufacturing the PS3 together, so perhaps we could glean a clearer picture of the price tiers Sony deciderd to go with concerning their sku's.

So we will need the following information, and any and all constructive comments and discusion is welcome:

*The cost of manufacturing the cell chip as of this moment.
*The cost of manufacturing RSX as of this moment.
*The cost of the memory as of this moment
*The cost of manufacturing a blu-ray disc drive as of this moment
*The cost of both a 7200 RPM 60gig and 20gig 2.5 inch hard disc drive
*And the cost of all other components such as Wifi, USB, Bluetooth, I/O, including HDMI, and the shell itself
*plus the cost of the wireless controler

Now it's important to remember that these will all be produced in bulk which will lower the cost of each... most likely 6 million units of each item above.

So If anyone has any information concerning any of the above, please feel free to share.

Here are my best guess estimates to start us off (by no means accurate, and based on the Meril Lynch report on the cost of manufacturing the PS3):

Cell = 200
RSX = 70
Blu-ray = 240
Memory = 50
HDD 20 gig = 26
HDD 60 gig = 40
USB Ports, wifi, and bluetooth = 15
other components = 65
Shell = 10
Controller = 25

So my total comes to 715 dollars for the premium pack

for some things above I deducted 33 percent from Meril Lynch's reportings so as to account for mass producing... the HDD, the Cell, Blu-Ray drive and the other components.

Feel free to dissagree with or to add to the above.

here is a link to the Meril Lynch report: http://rsch1.ml.com/9093/24013/ds/276873_0.PDF

 

[nintenho]

How the hell could you find out? Some said that the PS2 cost 700 to make at launch.

 

[MBDF]

How the hell could you find out? Some said that the PS2 cost 700 to make at launch.

Well, some people here are very knowledgeable concering things like this, so I figured it was worth a try.

 

[nintenho]

Well, some people here are very knowledgeable concering things like this, so I figured it was worth a try.

But you need very specific information. And Sony is probably selling it at a loss so what are you gaining from this thread?

 

[Guden Oden]

Here are my best guess estimates to start us off

Feel free to dissagree with or to add to the above.

Um, since you're just guessing, what's the point?

By the way, you put 20 bucks for USB and wifi but only 5 for the very large and shiny chrome-detailed casing with piezo touch buttons and stuff, I wouldn't be surprised if it isn't more like the other way around.

This thread doesn't sound very technical to me by the way, or even useful. There's no way any of us could get accurate information for this anyway as it's all trade secrets. Besides, why should we even care?

 

[Shifty Geezer]

Just search for the Analyst BOM prices thread where there's discussion, and you'll see no-one has any knowledge, really. Guesses can range from $600 - $1000 depending on total pie-in-the-sky figures. With no idea of Cell yields for example, it's impossible to pin a price on the Cell processor used in PS3.

 

[_xxx_]

$1000 is really too much, the price should be anywhere between $400-600. No way is it more than $700, worst-case.

Bear in mind that mass produced components are cheaper, so calculating the price by adding individual estimated parts' prices makes no sense.

 

[Urian]

In the book named "Computer Architectures: A quantitative approach" by the doctors Hennessy and Patterson exist a table that says that a processor with Cell size made in a 150mm^2 waffer has a cost of 199$.

But Cell is made in a 300mm^2 waffer and this reduces a lot the cost, but this can help you with the prediction because Cell cannot be more expansive than 199$ unless the yields are very bad.

 

[Shifty Geezer]

Many would agree at the $1000 price, but some were pegging it that high. I don't think $500 is anything like reasonable though, otherwise Sony are going for break-even this time around, at a risk of losing market share. If we assume they're making a loss on the hardware as usual, price would be $600-$700 for the entry level PS3, no? But then, maybe they are wanting to make money from the hardware early on this time?

 

[MBDF]

In the book named "Computer Architectures: A quantitative approach" by the doctors Hennessy and Patterson exist a table that says that a processor with Cell size made in a 150mm^2 waffer has a cost of 199$.

But Cell is made in a 300mm^2 waffer and this reduces a lot the cost, but this can help you with the prediction because Cell cannot be more expansive than 199$ unless the yields are very bad.

This is exactly the kind of info I was hoping for, interesting. I'll update my prediction above.

Thanks Urian.

 

[Npl]

It costs 666$ to make. To survive the devil you gotta become the devil

 

[pakotlar]

It costs 666$ to make. To survive the devil you gotta become the devil

It's actually $999. Sony used a reflection to fool the world into believing the devil never existed. Sony's head is curiously nicknamed "reficul".

 

[TheChefO]

It's actually $999. Sony used a reflection to fool the world into believing the devil never existed. Sony's head is curiously nicknamed "reficul".

be prepared

 

[Urian]

Are you making material for Deadmeat habitual posts in Opa-Ages?

 

[Carl B]

Obviously a clear answer is beyond us here as there are just too many unknowns. But Sony themselves provided some of the most relevant information to date recently on their year-end conference call.

If we take the ~$850 loss they are predicting through the fiscal year and the 6 million consoles shipped that figure hinges upon, then...

(x * 6 million) + y = z

where:

x = the amount lost per console

y = estimated profits from continuing PS2 and PSP operations

z = the (~$850) million estimated loss for SCE in the present fiscal year

Now I have no idea what profits sans PS3 launch would look like this year, but if we just arbitrarily put it at $300 million (which seems normal enough for SCE), then the loss per console would average about $191 across both SKUs and all territories.

Personally I think a profit of $300 million otherwise might be too high though near the end of PS2's heyday, and the losses associated with PS3 at this stage would include expenditures (such as on marketing) seperate from the console manufacturing itself. So for myself I'm guessing at $150-175 lost per console across both SKUs and all territories. And it's important to add that in of course, because one of these SKUs - we don't know which - will be costing Sony more, and different territories will be costing Sony more as well, US probably being the 'lossiest.' All we can determine through the numbers we have is the average.

 

[Urian]

Obviously a clear answer is beyond us here as there are just too many unknowns. But Sony themselves provided some of the most relevant information to date recently on their year-end conference call.

If we take the ~$850 loss they are predicting through the fiscal year and the 6 million consoles shipped that figure hinges upon, then...

(x * 6 million) + y = z

where:

x = the amount lost per console

y = estimated profits from continuing PS2 and PSP operations

z = the (~$850) million estimated loss for SCE in the present fiscal year

Now I have no idea what profits sans PS3 launch would look like this year, but if we just arbitrarily put it at $300 million (which seems normal enough for SCE), then the loss per console would average about $191 across both SKUs and all territories.

Personally I think a profit of $300 million otherwise might be too high though near the end of PS2's heyday, and the losses associated with PS3 at this stage would include expenditures (such as on marketing) seperate from the console manufacturing itself. So for myself I'm guessing at $150-175 lost per console across both SKUs and all territories. And it's important to add that in of course, because one of these SKUs - we don't know which - will be costing Sony more, and different territories will be costing Sony more as well, US probably being the 'lossiest.' All we can determine through the numbers we have is the average.

I prefer to watch to the size of the processors and memory chips inside the system and they are very similar to the 360 processors in size, without the BD-ROM Sony coul sell PS3 for 399$ losing 125-150$ and I don´t believe that the BluRay Drive is going to cost 200$. I am sure that Sony loses less money with the 600$ pack than MS with the 400$ pack.

 

[Bob]

In the book named "Computer Architectures: A quantitative approach" by the doctors Hennessy and Patterson exist a table that says that a processor with Cell size made in a 150mm^2 waffer has a cost of 199$.

I assume you meant a wafer that's 150 mm in diameter. 150 mm^2 is smaller than Cell in 90 nm.

In either case, I'm not sure where you get that figure. What kind of defect density are you assuming? Do you know how well that matches up with Cell's actual defect density? What kind of complexity factor (alpha) are you using? If Cell is built with 8 SPEs but one can be turned off to increase the number of functional chips, how does that affect yields and/or price?

Lots of assumptions there...

 

[Carl B]

I prefer to watch to the size of the processors and memory chips inside the system and they are very similar to the 360 processors in size, without the BD-ROM Sony coul sell PS3 for 399$ losing 125-150$ and I don´t believe that the BluRay Drive is going to cost 200$. I am sure that Sony loses less money with the 600$ pack than MS with the 400$ pack.

Sure, I would prefer that also - but since we don't have hard figures on any of those, we have nowhere to go with it really. I mean I personally think that Cell and RSX should cost less in PS3 than EE and GS cost in the launch PS2; not only are they smaller than those two and on larger wafers, but the process is more mature than the 250nm node Sony launched on however many years ago, and the chip has actually been in production for quite some time now. So the expense is coming from somewhere, and I think BD is the biggest culprit. Thankfully it will also be something that reduces in cost fairly rapidly once a concerted industry-wide BD effort begins.

But without any of those numbers, for myself I think the financials route is the way to go. Sony knows their own costs, and they are estimating a $850 loss on 6 million consoles. Works for me!

 

[nelg]

When did R & D become free?

 

[Carl B]

When did R & D become free?

R&D doesn't factor into the cost to build though per console - R&D has been paid on PS3-related research ever since 2000 I imagine, and it is expensed as a normal course of business. Basically, PS2 operations pay for PS3 research, PS3 operations (and profits) will pay for PS4 R&D over the next five/six years, and so on...

R&D isn't free, but it's already been paid. Some people have the notion in their heads that R&D costs should be divided across the total number of consoles produced. That would be an interesting post-morten figure to derive once production on said product has halted, but for the purposes of the thread question, R&D isn't very relevant.

Now, if what you're saying Nelg though is that we should seperate out any would-be R&D costs from the $850 million loss for the year before calculating cost-per-console, I agree with you, but we just don't have those figures.