General Next Generation Rumors and Discussions [Post GDC 2020]

[iroboto]

https://twitter.com/x/status/1283506875976351745

Yields are bad but can double production. Only on B3D.

edit: sorry, increase it by 50%. Don't want to get called out.

Production amounts and yields aren’t the same thing.

How much you decide to produce has to do with market demand. Yield is only a matter of the number of useable chips available for production. They aren’t the same. If chip A gets 85% yield and another 90%, that doesn’t mean you’re doing to produce 5% less (well you are producing less usable chips). But the end result means you’re going to pay more to produce the same amount usable chips. Whether they want that to be reflected in price to the customer is up to Sony.

Sony could easily produce 30M chips if they wanted to. Doesn’t mean they can assemble and ship that many.

the issue here comes down to how much you’re willing to spend on your first production run. There are all sorts of parts that need to come together in the assembly pipeline for this to happen. So if the rest of the suppliers can supply the parts ramping silicon production is relatively straight forward. You also don’t want warehouses full of product sitting around as that leads to extra cost, so you want to produce as much as it’s selling otherwise you’ve got a problem.

Chip yield, power consumption, heat dissipation are all solved; Sony won’t release a product that will have issues, they will just engineer the solution for it; It just gets reflected in cost, which they will need to choose whether to pass that down to price.

 

[Deleted member 7537]

Companies like intel having issues because they cannot meet demand, but Sony can produce as many APUs as they want to make up for bad yields. Got it.

 

[Barrabas]

Lucky for Sony they don't have 'fixed clocks' to lower this gen

No, they have variable clocks to adjust depending on yields....

 

[Deleted member 11852]

Production amounts and yields aren’t the same thing. How much you decide to produce has to do with market demand

Very true and to make matters more complicated, production and ship timeframes from Asia to anywhere other than Asia (i.e. the Americas, Europe, Middle East, Africa) is 2-3 months so you're actually producing to what demand you expect in 8-12 weeks. It's desirable to more production capability than demand because you don't want to run out of consoles which is something that Nintendo still seem to struggle with.

But if you're consistently producing more than you're selling and a) you know you're not imminently price-dropping to fuel sales, and b) there are no highly anticipated games releasing soon to fuel sales, then you can slow production a little. Semiconductors are not an issue to store so you'll slow console production rather than semiconductor production - you can fit a lot of ICs on a standard 1200x1000 pallet stacked 4-5ft high.

It's surprising and unexpected to see Sony pivot from April where they were expecting to ship far fewer consoles than PS4's launch to now where they're looking at around 25% more. Official pre-orders have not begun yet so what is driving this confidence of high demand in the middle of the coronavirus-sponsored 2020?

 

[Shifty Geezer]

Yields change. Good yields now don't mean there wasn't bad yields before.

Curious what could change though. The production is the same as XBSX. The architecture is effectively the same. If we assume there were bad yields as not enough chips were able to clock high enough, what can be done to fix that before a node shrink? Can TSMC tweak the fab process somehow?

Production amounts and yields aren’t the same thing.

Yes, but if yields are low, price is higher. The earlier rumour was fewer PS5s, no? This was presented as quite a problem. So either that's true and Sony are spending a lot of money to make enough workable chips, with lots of duds going in the bin, or they are going to charge lots for the console to cover those losses.

The general narrative based on rumours without evidence has been that Sony has production difficulties due to their architecture. Here we have better sources suggesting no production issues. Deciding whether the rumours were true but things have changed, versus the rumours were never true to begin with, seems to come down to personal preferences.

 

[iroboto]

Companies like intel having issues because they cannot meet demand, but Sony can produce as many APUs as they want to make up for bad yields. Got it.

Intel doesn’t run out of celerons and low end pentiums. It’s their high end core i7/i9 and above where the price points for an individual CPU are 2-3x the cost of a PS5.

yields between a low clocked 300mm^2 low performance with added redundancy vs 450+mm for cpu alone no redundancy and super high clocks made to run in extreme conditions is not an equal comparison.

couple with intel being responsible for their own fabrication and the fact that intel owns a sizeable market of all computing, then its quite believable that they would run into demand issues if all of their products are in high demand.

 

[goonergaz]

Very true and to make matters more complicated, production and ship timeframes from Asia to anywhere other than Asia (i.e. the Americas, Europe, Middle East, Africa) is 2-3 months so you're actually producing to what demand you expect in 8-12 weeks. It's desirable to more production capability than demand because you don't want to run out of consoles which is something that Nintendo still seem to struggle with.

But if you're consistently producing more than you're selling and a) you know you're not imminently price-dropping to fuel sales, and b) there are no highly anticipated games releasing soon to fuel sales, then you can slow production a little. Semiconductors are not an issue to store so you'll slot console production rather that semiconductor production - you can fit a lot of ICs on a standard 1200x1000 pallet stacked 4-5ft high.

It's expected to see Sony pivot from April where they were expecting to ship far fewer consoles than PS4's launch to now where they're looking at around 25% more. Official pre-orders have not begun yet so what is driving this confidence of high demand in the middle of the coronavirus-sponsored 2020?

maybe surveys are giving them confidence in the product? Maybe it will be cheaper than the general thought out there? Maybe they will offer a trade in?

Curious what could change though. The production is the same as XBSX. The architecture is effectively the same. If we assume there were bad yields as not enough chips were able to clock high enough, what can be done to fix that before a node shrink? Can TSMC tweak the fab process somehow?

Yes, but if yields are low, price is higher. The earlier rumour was fewer PS5s, no? This was presented as quite a problem. So either that's true and Sony are spending a lot of money to make enough workable chips, with lots of duds going in the bin, or they are going to charge lots for the console to cover those losses.

The general narrative based on rumours without evidence has been that Sony has production difficulties due to their architecture. Here we have better sources suggesting no production issues. Deciding whether the rumours were true but things have changed, versus the rumours were never true to begin with, seems to come down to personal preferences.

I believe there is a segment of spin doctors out there - first it was how loud PS5 was, then expensive cooling, then poor production levels - this all goes along with the anti variable frequency noises - anything to disrupt or play down the opponent...it kind of reminds me of Alex Ferguson and his mind games.

 

[Deleted member 11852]

maybe surveys are giving them confidence in the product? Maybe it will be cheaper than the general thought out there? Maybe they will offer a trade in?

There are lies, damned lies and survey results.

Sony clearly have reason to be almost insanely bullishly confident. I guess we'll find out what that is in due course.

A bit O/T for this thread but I believe that Sony are willing to eat a bigger loss on the first year's console sales and I think that Sony expect to sell more consoles than Microsoft in both the short and long-term with part of their strategy being that PS5-exclusive games (Ratchet & Clank, Spider-Man etc) around launch will give PS5 an appeal-edge over Series X in that first year. The enemy is price but in the long term, looking at Sony's revenue from PSN and other services, they will dwarf potential loses if they can get these people on board and spending on services for longer.

edit: Microsoft have today confirmed that there will be no Xbox Series X exclusives from MGS for the "next couple of years".

 

[Deleted member 7537]

Intel doesn’t run out of celerons and low end pentiums.

https://embedtek.net/knowledge/intel-cpu-shortage-continues/

Intel confirmed formally that which we have been aware of for several weeks – the shortage of Intel CPUs, specifically lower-end Celeron, Pentium and I3 series processors will persist through the first half of 2020.

https://www.techadvisory.org/2019/02/expect-cpu-shortages-until-late-2019/

Businesses who are unable to acquire the newest Core, Celeron, and Pentium series processors for desktops and laptops may have to look for alternatives compatible with non-Intel chips.

Anyway, the point is some rumors (amplified by "journalists" at Windows Central) came out some weeks ago about Sony having serious issues with yields. Rumors which were never confirmed. Now some respectable outlets like Nikkei or Bloomberg, which may be wrong, talk about a 50-100% increase in production. Too keep selling the story of how Sony completely dropped the ball with the PS5, panicked and increased the frequency of the chips beyond what was reasonable, now the argument is that Sony has magically increased production enough to produce 10 million working PS5 APUs and throw to the garbage the ones that don't.

 

[chris1515]

There are lies, damned lies and survey results.

Sony clearly have reason to be almost insanely bullishly confident. I guess we'll find out what that is in due course.

A bit O/T for this thread but I believe that Sony are willing to eat a bigger loss on the first year's console sales and I think that Sony expect to sell more consoles than Microsoft in both the short and long-term with part of their strategy being that PS5-exclusive games (Ratchet & Clank, Spider-Man etc) around launch will give PS5 an appeal-edge over Series X in that first year. The enemy is price but in the long term, looking at Sony's revenue from PSN and other services, they will dwarf potential loses if they can get these people on board and spending on services for longer.

The first part of the PS4 lifetime from 2013 to 2016 was not very good for exclusives games. Andrew House told it in 2015 and Shuhei Yoshida said in 2018 or 2019. They were happy of first party exclusive release in quantity and quality from 2017.

They will sell more games and they will recover the loss. All first party games are for the PS5 first year. And this time digital sales are higher than during first year of current-gen.

And never forget, Jim Ryan told this to investor and shareholders:

As we move towards the next-generation in 2020, one of our tasks – probably our main task – is to take that community and transition it from PlayStation 4 to PlayStation 5, and at a scale and pace that we’ve never delivered on before.”

 

[iroboto]

https://embedtek.net/knowledge/intel-cpu-shortage-continues/



https://www.techadvisory.org/2019/02/expect-cpu-shortages-until-late-2019/



Anyway, the point is some rumors (amplified by "journalists" at Windows Central) came out some weeks ago about Sony having serious issues with yields. Rumors which were never confirmed. Now some respectable outlets like Nikkei or Bloomberg, which may be wrong, talk about a 50-100% increase in production. Too keep selling the story of how Sony completely dropped the ball with the PS5, panicked and increased the frequency of the chips beyond what was reasonable, now the argument is that Sony has magically increased production enough to produce 10 million working PS5 APUs and throw to the garbage the ones that don't.

In 2019 intel had fallen in shortage of their server high end CPU product. They opted to swap out their low end products in favour to serve their high end products. supply issues for a company with the scope, scale and reach of intel makes sense especially when they a responsible for fabrication such a large wide variety of products, the two are not equivalent. I just wanted to point out drawing a comparison of a single SKU APU vs the whole company product line is not a equivalence of comparison.

Yield issues for Sony would manifest in cost to the company. Poorer yield would reflect in cost per chip. At a $1 more per chip, or $5 more per chip you are looking at a lifetime of 500M in cost over the generation. At $10 per chip you’re looking at 1B unaccounted for.

 

[Deleted member 11852]

Anyway, the point is some rumors (amplified by "journalists" at Windows Central) came out some weeks ago about Sony having serious issues with yields. Rumors which were never confirmed. Now some respectable outlets like Nikkei or Bloomberg, which may be wrong, talk about a 50-100% increase in production.

Information like disappointing yields can be blown out of proportion and whether something did good is bad is tied to expectation.

If Sony/AMD/TSMC expected 90% yields but only got 87% they might be disappointed.

 

[iroboto]

Information like disappointing yields can be blown out of proportion and whether something did good is bad is tied to expectation.

If Sony/AMD/TSMC expected 90% yields but only got 87% they might be disappointed.

I think people immediately jump to poor yields as being 30-50% as that is how we look at most things. When something as simple as 1-2% poorer yield over 100M devices will add up over life time in cost; that 100M needs to come out of somewhere, that is the definition of poor yield for a low end consumer device.

 

[manux]

Could we just dismiss the poor yields? All evidence points otherwise other than something that could easily just be console war:ing/marketing FUD.

If PS5 in general is loud and sucks tons of power despite large case then perhaps there is something to sony pushing the chips too hard via "overclocking". We wont know this until launch. In the spirit of FUD all the non release consoles that press etc. will see before launch could be cherry-picked.

 

[Deleted member 11852]

I think people immediately jump to poor yields as being 30-50% as that is how we look at most things. When something as simple as 1-2% poorer yield over 100M devices will add up over life time in cost; that 100M needs to come out of somewhere, that is the definition of poor yield for a low end consumer device.

You would expect the yields to improve over time so it should not be an issue for 100m chips. If it is, AMD and/or TSMC screwed up.

 

[iroboto]

You would expect the yields to improve over time so it should not be an issue for 100m chips. If it is, AMD and/or TSMC screwed up.

I generally do expect things to get better. Node changes and major revisions etc.

 

[iroboto]

Could we just dismiss the poor yields? All evidence points otherwise other than something that could easily just be console war:ing/marketing FUD.

If PS5 in general is loud and sucks tons of power despite large case then perhaps there is something to sony pushing the chips too hard via "overclocking". We wont know this until launch. In the spirit of FUD all the non release consoles that press etc. will see before launch could be cherry-picked.

Why would poor yields or power issues result in a loud fans ?

 

[manux]

Why would poor yields or power issues result in a loud fans ?

There was some mentions here and elsewhere that sony compensated "poor yields" by pushing more power to the chips to achieve the clock speeds. i.e. bad yields being that using sensible voltage the chips would not reach their specified clock speed. In my books this is FUD. Sony would have known the characteristics of their chips for long time now and it wouldn't be a surprise towards end of the project.

edit. PS5 soc is not that big of a chip presumably. It would be unlikely it would have a lot of defects and bad yields in general. AMD has enough experience with 7nm to make that a non issue. Or at least it should not be yielding any worse than any other amd designed 7nm product. The open item in the fud is achieving the 2.3GHz clock speed and chips not yielding that high clock speed.

 

[Shifty Geezer]

I believe there is a segment of spin doctors out there - first it was how loud PS5 was, then expensive cooling, then poor production levels - this all goes along with the anti variable frequency noises - anything to disrupt or play down the opponent...

Please keep console warring out of it. Some people might always have a particular interpretative leaning, but as long as the arguments are fairly and civilly presented, it's all good.

 

[Shifty Geezer]

I think people immediately jump to poor yields as being 30-50% as that is how we look at most things.

That's certainly the way the rumours had presented things in this case. It was presented as a significant problem in PS5's design (the high clock speeds) with notable impact on availability and/or costs.

If the original info was just, "Sony not seeing the yields they hoped for," responses and arguments may have been more conservative.