All purpose Sales and Sales Rumors and Anecdotes [2017 Edition]

Status
Not open for further replies.
I don't think realistically the consoles can drop much further in price.
As I understand it, Moore's Law wasn't about just doubling transistors every 2 years, but doubling in which it was economically feasible.

Smaller and smaller nodes require a much larger process to manufacture. And Moore's law ended at 28nm where the price was low to manufacture and could over time continue to be cheaper.

I don't see 16nm getting all that much cheaper the next 3 years. 7nm will definitely be more expensive. Nvidia charts show wafer costs as being the highest at 7nm.

That leaves memory and hard drive prices as a deferment factor for price drops and that situation could take a while to clear up.

Observe spike in costs going from 28 to 14
NV-Pres4.jpg


NV-Pres3.jpg


The move to 7nm should be significantly higher I imagine.

The slim variants launched Q3-Q4 2016 with Scorpio in Q4 2017. Not sure how much further that graph will go down.
I assume 16nm is cheaper than 14nm though.

If you look at 2nd graph, the move from 28nm to 14nm the normalized cost per transistor is very close. Except we doubled the number of transistors, so we're paying double. I don't see prices dropping much further on the silicon.

It almost seems impossible to see another $399 console without waiting a long time imo. If these normalized costs are correct and this pattern continues. We should see the companies strike at the point in which the normalized transistor cost is better than 16nm at that point in time.

Quite a bit more transistors however.

From extreme tech:
Cortex_A17_1.jpg


I think from the 3 graphs together, drawn in many different ways, we can only see more expensive consoles go forward (if we stay on the current APU path) until there is a technology switch. But that also means we are unlikely to see much more price drops as a result of wafer costs, so we're really just looking at memory and HDD prices to dictate the next set of price decreases.

Imo, the probabilities that PS4 Pro will ever get to $199, with the way things are today, are extremely low. This may also shed a bit more light on the importance of the mid gen refresh.
 
Last edited:
Base PS4 is not going anywhere. It will remain in sales well after PS5 appears.

Maybe. But look at it from an Apple model type perspective. Let’s say that the PS4 is the iPhone 4 and the Pro is the 4 S ...
 
Maybe. But look at it from an Apple model type perspective. Let’s say that the PS4 is the iPhone 4 and the Pro is the 4 S ...

Sony traditionally continue to manufacture (and refine designs) of consoles long after they've been superseded. PS2 launched is 2001, PSone wasn't discontinued until 2005. PS3 launched 2006/07, PS2 wasn't discontinued until 2013! PS4 launched 2013, Sony finally discontinued PS3 this year.
 
I don't think realistically the consoles can drop much further in price.
As I understand it, Moore's Law wasn't about just doubling transistors every 2 years, but doubling in which it was economically feasible.

Smaller and smaller nodes require a much larger process to manufacture. And Moore's law ended at 28nm where the price was low to manufacture and could over time continue to be cheaper.

I don't see 16nm getting all that much cheaper the next 3 years. 7nm will definitely be more expensive. Nvidia charts show wafer costs as being the highest at 7nm.

That leaves memory and hard drive prices as a deferment factor for price drops and that situation could take a while to clear up.

Observe spike in costs going from 28 to 14
NV-Pres4.jpg


NV-Pres3.jpg


The move to 7nm should be significantly higher I imagine.

The slim variants launched Q3-Q4 2016 with Scorpio in Q4 2017. Not sure how much further that graph will go down.
I assume 16nm is cheaper than 14nm though.

If you look at 2nd graph, the move from 28nm to 14nm the normalized cost per transistor is very close. Except we doubled the number of transistors, so we're paying double. I don't see prices dropping much further on the silicon.

It almost seems impossible to see another $399 console without waiting a long time imo. If these normalized costs are correct and this pattern continues. We should see the companies strike at the point in which the normalized transistor cost is better than 16nm at that point in time.

Quite a bit more transistors however.

From extreme tech:
Cortex_A17_1.jpg


I think from the 3 graphs together, drawn in many different ways, we can only see more expensive consoles go forward (if we stay on the current APU path) until there is a technology switch. But that also means we are unlikely to see much more price drops as a result of wafer costs, so we're really just looking at memory and HDD prices to dictate the next set of price decreases.

Imo, the probabilities that PS4 Pro will ever get to $199, with the way things are today, are extremely low. This may also shed a bit more light on the importance of the mid gen refresh.
I'm beginning to think there is something either wrong or missing from nvidia graphs.

Between 28nm and 14nm the PS4 went from $399 to $249. It's even $199 for BF. A similar graph above was used in 2013 to say the PS4 wouldn't drop in price and it certainly did.
 
I'm beginning to think there is something either wrong or missing from nvidia graphs.

Between 28nm and 14nm the PS4 went from $399 to $249. It's even $199 for BF. A similar graph above was used in 2013 to say the PS4 wouldn't drop in price and it certainly did.

How big part of ps4 cost is due to 28nm soc? How much did 28nm mature during ps4 lifetime driving soc cost down?
 
I'm beginning to think there is something either wrong or missing from nvidia graphs.

Between 28nm and 14nm the PS4 went from $399 to $249. It's even $199 for BF. A similar graph above was used in 2013 to say the PS4 wouldn't drop in price and it certainly did.
The price per transistor dropped by 2016 below the cost of 28nm. But the number of transistors did not increase, thus we should see a reduction in price.
 
Sony traditionally continue to manufacture (and refine designs) of consoles long after they've been superseded. PS2 launched is 2001, PSone wasn't discontinued until 2005. PS3 launched 2006/07, PS2 wasn't discontinued until 2013! PS4 launched 2013, Sony finally discontinued PS3 this year.

Each time Sony had 2 consoles being manufactured and on the market, you're suggesting they will have 3 when PS5 launches and won't be in a hurry to drop one? I guess the sales will dictate what Sony end up doing (if they're making money from PS4 why drop it?) but I suspect they will get a bit of a stock pile of PS4s and then sell them out and Pro will because the new entry level PlayStation @ ~ £200

I don't think they will want to juggle 3 consoles for too long unless there's good money to be made and it's not damaging PS5 sales.
 
Naturally as this generation lives out, the longer it lives the more the ratio should in theory move towards 4Pro; I'll nod to @Globalisateur point there; however the ratio of change should fall in line with the purchases of 4K TV sets, and not some random number.

transitions take a lot of time! What's the point of releasing PS5 with specs to ring in 4K native next generation graphics, when no one has it? That's just a super waste of power, you've sent a console to die because you delivered a solution for where no problem exists. We need to wait until the market has transitioned to 4K before releasing a proper successor. We may need to wait until 7nm and see the price come down on 4Pro and 1X with the reduction as well as the memory price reduction.

Think about all the other zones that won't have 4K for a long time! 4K is still very much a NA thing.
 
Wouldn't PS4 be better in that respect than PS4Pro, because it'll be a larger gap for PS5 and make PS5 look that much better as an alternative at the higher price range?

Possibly, but PS5 will (IMHO) have exclusives and be a true fully native 4K60 machine (assumed) and this will (along with playing cross-gen games noticeably better ) be enough to justify the upgrade.

It didn’t work for the PS3.... :D

? PS3 may have been considered the most powerful but 360 won pretty much all face-offs yet once launched the PS3 outsold the 360 closing the head-start until ending on near equal sales?

Naturally as this generation lives out, the longer it lives the more the ratio should in theory move towards 4Pro; I'll nod to @Globalisateur point there; however the ratio of change should fall in line with the purchases of 4K TV sets, and not some random number.

transitions take a lot of time! What's the point of releasing PS5 with specs to ring in 4K native next generation graphics, when no one has it? That's just a super waste of power, you've sent a console to die because you delivered a solution for where no problem exists. We need to wait until the market has transitioned to 4K before releasing a proper successor. We may need to wait until 7nm and see the price come down on 4Pro and 1X with the reduction as well as the memory price reduction.

Think about all the other zones that won't have 4K for a long time! 4K is still very much a NA thing.

I kind of agree but when you say 'very much a NA thing' don't forget the UK. Those are 2 of the most important markets making up around 50% of total sales AFAIK.
 
I kind of agree but when you say 'very much a NA thing' don't forget the UK. Those are 2 of the most important markets making up around 50% of total sales AFAIK.
Absolutely, I was under the impression 4K TV sales were not as big there as here yet, (even then NA are just starting to ramp up now).

I do believe that these consoles will drive TV set sales. By how much I'm unsure. But if UHD, and streaming content is slim or the markets haven't grown enough yet, then games would be the natural remaining 4k entertainment device capable of fully leveraging the technology. In some ways even more so since games don't suffer from video compression on output.
 
? PS3 may have been considered the most powerful but 360 won pretty much all face-offs yet once launched the PS3 outsold the 360 closing the head-start until ending on near equal sales?
That may or may not have been due to being perceived 'most powerful' - other factors like brand also played a role. In other examples of a truly most powerful console, like XBox or NeoGeo, being most powerful wasn't as important for sales as the rivals being most popular and most fashionable and cheaper, etc.
 
I'm beginning to think there is something either wrong or missing from nvidia graphs.

Between 28nm and 14nm the PS4 went from $399 to $249. It's even $199 for BF. A similar graph above was used in 2013 to say the PS4 wouldn't drop in price and it certainly did.
True, but I just realised PS4/XO console prices aren't the best indicator of wafer/fab costs. XO/PS4 afterall build their business model on their 30% cut of the digital distribution platform, their cut on post sale dlc and microtransactions, software royalties, and $50-60 PLUS/Gold membership, and 1st party software revenue.
So they can absorb price fluctuations in component costs and even sell at losses, unlike EVGA who have to make their margins off of hardware.

Also I've read competition between fabs doing high performance chips (hot chips) is extremely stiff and margins are low. We've gone from 20 companies doing that kind of fabrication 20 years ago to only 4-5 nowdays. I can't remember who they mentioned but another one will be likely to exit in a few years as well. Did they mention IBM or Global Foundries? I can't remember.
 
Last edited:
Absolutely, I was under the impression 4K TV sales were not as big there as here yet, (even then NA are just starting to ramp up now).

Yes, but you were talking about PS5 which is at least 2 years away - by then 4K will be far more commonplace. I don't believe people (will) want 4K like they did 1080p for the resolution upgrade...I think a lot of sales will be on the back of HDR, for me that's the easiest sell. Obviously that depends on content so it might be a wash...I just feel that 4k vs 1080p is diminishing returns.
 
Status
Not open for further replies.
Back
Top