D
Deleted member 11852
Guest
Can we take the AMD vs Nvidia to a more appropriate thread.
It's irrelevant here.
It's irrelevant here.
PS4 Pro Speculation (PS4K NEO Kaio-Ken-Kutaragi-Kaz Neo-san)
- Thread starter mpg1
- Start date
-
- Tags
- sony sony roach motel
- Status
To be honest, it's the entire PC vs console tangent that is irrelevant.
Assuming a shrink of 2x, the 28nm PS4 APU would drop to ~175mm2, which is smaller than the smallest AMD GPU with a 256-bit GDDR5 bus, Pitcairn.
I'm not sure where the shrink could become pad-limited by the external interfacing requirements for the current memory and IO pads. That would put a floor in what a shrink could do to save money before it's just paying for blank area. Doubling the GPU would probably push things back past that point.
Slims should have higher volume requirements coupled with lower prices, which Sony probably doesn't want to dedicate larger silicon chips that should be going into more expensive boxes more than it has to. Neo's yields would have to be notably poor in order to satisfy that segment.
However, if there's a business case for a shrunken PS4 chip, then mixing real slims and cut-down Neo chips might be a hassle.
When Mark Cerny called it a "super-charged PC architecture" it's evident that PS4 is essentially a PC with fewer kinds of models and a closed architecture just like Mac. With x86, unlike Cell, they don't have to think about the long term ventures like depreciation of semiconductor factories by themselves. What they have to think about is how they can maximize profit and reach at the global level, some countries need cheaper, affordable models, other countries need pricier and shinier models, etc. The variations with different HDD size could justify the prices of more expensive models, why not try more substantial customization with the additional bonus of UHD BD & HDR support?
Interesting take: Sony and MS have to find a way to create a premium price model to maximize revenues and since PS4 and Xbox One can be upgraded with inexpensive 3rd-party drives (internal, external) that is no longer a viable way to market a premium unit.
http://venturebeat.com/2016/04/20/why-sony-needs-the-playstation-4-neo-even-if-you-hate-it/
http://venturebeat.com/2016/04/20/why-sony-needs-the-playstation-4-neo-even-if-you-hate-it/
DuckThor Evil
Legend
Other than launch numbers. Which records have been broken? PS4 is doing very well, but it's not beating Wii's numbers and not really even PS2 when launches are aligned. Xbox One isn't really setting the world on fire and neither is the Wii U.
Can you tell by the die shot from chipworks?
D
Deleted member 11852
Guest
PS4 outsold PS2 (launches aligned) last year. I'm talking a combination of consoles sales, game sales and revenue. Wii sold console units but not software and you need software sales, particularly third party software sales, for a vibrant ecosystem. Otherwise what you have is a Wii or Wii U
Never have gamers been so quick to abandon the last gen and spend on the new gen.One dimension of the chip is effectively taken up by half the GDDR5 bus. The other dimension has a little over half taken up by the interface.
A naive shrink of the overall chip area could still happen, although if some claims of greater than 2x were to occur an unchanged PHY configuration is going to leave a rectangle larger than the silicon would need.
The Xbox One's interface doubles up the PHY, so overall perimeter is less of an issue there. I don't recall an example of that for GDDR5, however.
hum... Design in some dead space (bonus thermal spacing)? Guess it'd still be waste even if they were to pad things up by introducing more redundant blocks (not sure how beyond the compute units).
oof.
hm...
DuckThor Evil
Legend
PS4 outsold PS2 (launches aligned) last year. I'm talking a combination of consoles sales, game sales and revenue. Wii sold console units but not software and you need software sales, particularly third party software sales, for a vibrant ecosystem. Otherwise what you have is a Wii or Wii U
The link you gave presents a situation that is not even close to being launch aligned. Instead it only has little over one year of NA and Europe sales of PS2. When you actually aling the launches properly. PS2's numbers are much better. This gen is doing alright...
DuckThor Evil
Legend
It's probably cheaper to use the old 28nm process for the PS4 than go for 14nm at this point. 14nm sees a lot of demand from many players and is thus better suitable for bleeding edge designs. 28nm should have pretty good wafer pricing now when the demand is getting lower.
It can happen. Some of Intel's chips have space off in the corner they didn't really use, but that is partly due to having a modular architecture where it was a worthwhile tradeoff to reuse across multiple lines rather than hyper-optimize any specific product.
A custom APU might play more silicon Tetris since it doesn't have anything else to apply to.
The naive math I put forward is already pretty off since it does include the area taken up by the interfaces. Playing around with just trying to shrink the CPU and GPU blocks, it might be possible to fit quite a bit of hardware into a "slim" footprint since Sony may be constrained due to the GDDR5 perimeter and the display/USB/IO interface. Neo might be a case of modestly going beyond that limit.
Surely PS5 will be in the frame by the time 10nm and 7nm become viable.
PS5 will be 2018, just to piss people off.
Yes, but they would continue to sell the PS4K for many more years as the lower SKU, so the shrink is important to lower cost/power/size. The mid-gen release schedule is only enabled by a backward/forward compatibility, and it require this 2x overlap.
Just like the theory of a PS4 at 14nm in addition to the PS4K at 14nm launched at the same time.
DuckThor Evil
Legend
I wonder if it really is that important. The lower cost is questionable. New node tends to be more expensive and like discussed earlier, the dies are already small to the point where taking full advantage of the shrink is complicated. Also PS4 is already quite small. Does it really need to be smaller? PS4 Neo probably has around the same power draw if not slightly less. I think these consoles don't really rely on shrinking the die, until and if it really becomes much cheaper to do so and at that point we might be at a point where Sony already has 2 more powerful SKUs out.
It's an interesting question. Optimistically, and taking GF and AMD and Samsung claims at face value, you should be able to get about twice the performance for the same power by going wider or about 35% faster for the same power just by upping clock. I think that was mostly with reference to GPU, but I can't remember and I can't find the damn slides now!
Here's something from last year that alludes to something similiar: http://www.dailytech.com/TSMC+Hypes...ggles+to+Hit+Volume+at+16+nm/article37298.htm
Anway, it might be possible for Sony to get both the 30% clock jump for Jaguar and the double width GPU (with slight upclock) for something like close to the launch PS4. Perhaps they could even share the expensive HS&F and power modules with the PS4 if that was the case ...?
Edit: Or Neo could be based on a 500 mm^2 28nm chip and the system could be pulling 250W
- Status