Baseless Next Generation Rumors with no Technical Merits [post E3 2019, pre GDC 2020] [XBSX, PS5]

[Deleted member 13524]

To be fair, those don't contradict each other, but rather corroborate. Bloomberg didn't say they were AIMING at over $400, but that they ended up above that price accidentaly. So it only makes sence that they'd have a conservative SOC and design to reach the magical $399 price, eventhough they later end up with a $479 product because of unforseen rise in costs of certain parts.
Both point to to 399 TARGET.
I definetly don't trust any rummor 100%, but I'm just pointing out the Bloomberg narrative and the older github leak fit rather well together.

You're assuming the surprise aspect of the BoM amounts to no less than $80 over what they originally intended (which would be the same $370 as the OG PS4). That's a lot for a surprise.

What components would you believe was expected to cost X, but ended up being X+$80?
The article mentions NAND and RAM, but a quick visit to dramexchange will tell you there hasn't been any price hike on either front, on the contrary.

Could RAM and NAND still be more expensive than Sony predicted? Yes, but I doubt they'd miss the mark as far as $80.

I'll concede that maybe Sony was aiming for a $420-430 BoM so they could sell for $450 with almost no margins at first (like the PS4), and a $450 BoM puts them in a tough spot deciding whether to go up to $499 with clear profit or $449 with losses.

But there's nothing in that $450 BoM that says Sony was aiming at specifically $399, on the contrary.

 

[PSman1700]

It's your interpretation, not proof of anything.

Likewise yours that it aint the full chip, or target chip.

 

[Shifty Geezer]

Likewise yours that it aint the full chip, or target chip.

the difference is you are using your interpretation as proof of something, whereas people are trying to tell you that you have no proof, only possibility. Hence all your arguments about BOMs and whatever that hinge on this possibility cannot be discussed without discussing whether one believes your interpretation or not. Hence, fundamentally, all discussion relying on Bloomberg articles and GitHub leaks and insider info does not belong in the Tech Prediction thread. that thread is for people to look at real data like die sizes, thermal curves, market costs, market trends, etc., to make their own predictions.

All theories about rumours aligning belong elsewhere.

The only real documentation that should be discussed in the Tech thread is genuine leaks. The GitHub leaks themselves provide discussion worthy of the tech thread, but talk about their validity and scope in relation to rumours doesn't.

 

[mpg1]

You're assuming the surprise aspect of the BoM amounts to no less than $80 over what they originally intended (which would be the same $370 as the OG PS4). That's a lot for a surprise.

What components would you believe was expected to cost X, but ended up being X+$80?
The article mentions NAND and RAM, but a quick visit to dramexchange will tell you there hasn't been any price hike on either front, on the contrary.

Could RAM and NAND still be more expensive than Sony predicted? Yes, but I doubt they'd miss the mark as far as $80.

I'll concede that maybe Sony was aiming for a $420-430 BoM so they could sell for $450 with almost no margins at first (like the PS4), and a $450 BoM puts them in a tough spot deciding whether to go up to $499 with clear profit or $449 with losses.

But there's nothing in that $450 BoM that says Sony was aiming at specifically $399, on the contrary.

Maybe they were willing to eat it @ $399 if BOM had been around $420 but now can't...

 

[anexanhume]

Ah so a banned user, exited after hard data leaks occured.

Matt, another ’industry’ insider, well thats too bad then, again nothing. The problem is just that, insiders. Give us information thats relevant instead.



PS5 will be more powerfull then xbox, above 12TF. Journalists said so, atleast before bans etc

You must get tired moving that yardstick so much.

It doesn't proof anything, it's only doing regression tests for backwards compatibility with PS4 and PS4 Pro. That one line about being "full chip" might even mean something different to who wrote it, other than full PS5 chip. It's your interpretation, not proof of anything.

And what would be the point of going to the trouble of developing a 36CU chip to test it at 2Ghz if another one with more CUs and lower clock would be coming down the line still targeting 9TF? In your effort to insist with 9TF you start not making much sense.

The issue is that Oberon first appeared well before Navi was released. Rumors at the time suggested Navi had needed a design spin, so AMD engineers would have been working with release-candidate silicon at the same time. It makes sense that Navi 10 LITE would have been the only thing in usable form at the time for said regression testing. That was perfectly fine since no more than 36 CUs were needed for that testing. That configuration need not necessarily reflect final hardware.

 

[milk]

You're assuming the surprise aspect of the BoM amounts to no less than $80 over what they originally intended (which would be the same $370 as the OG PS4). That's a lot for a surprise.

What components would you believe was expected to cost X, but ended up being X+$80?
The article mentions NAND and RAM, but a quick visit to dramexchange will tell you there hasn't been any price hike on either front, on the contrary.

Could RAM and NAND still be more expensive than Sony predicted? Yes, but I doubt they'd miss the mark as far as $80.

I'll concede that maybe Sony was aiming for a $420-430 BoM so they could sell for $450 with almost no margins at first (like the PS4), and a $450 BoM puts them in a tough spot deciding whether to go up to $499 with clear profit or $449 with losses.

But there's nothing in that $450 BoM that says Sony was aiming at specifically $399, on the contrary.

Again, I am not betting any money on the Bloomberg article, I'm just pointing ou, sort of as a devil's advocate, that if fits in with the github leak story more than it does not.

But since you asked, and since shifty already moved this to the baseless thread here is how I interpret the Bloomberg article:

I feel that if it is not completely made up bullshit (and I think it very well could be) it is still a collection of rummors and hearsay that the journalist got access to, some through the same media stories we also saw, and some through "insider" gossip and what have you. So there would still be some logical leaps, personal interpretations and dot-connecting from the author. With articles like this, I don't take any specific word or number at face value, but rather I keep the general narrative as a possibility in the back of my head.

The general narrative of the bloomberg artiicle is: the ps5 design team was aiming at a cost and retail price close to that of the launch ps4 (maybe with some leeway)
their projections failed a bit. The cost is higher than they expected.
One of the excuses the team has been throwing around internally is increased cooling costs.
Another is unexpected memmory prices (blame it on those evil memory cartels and those pesky phone manufacturers, always a good scapegoat.)
Now, those excuses might hold some water, but they may not be the full story.

As for the memory cost surprise, another thing I had in the back of my head was how sony managed a last minute sweet deal for ps4's ddr4. Maybe they were sort of expecting the same bargaining power for ps5 in ther projections, but the industry shifted in such a way that suppliers are not as willing to cut as good a deal with sony this time as they were before.

And again, ALL of this can be conjecture from the journalist. Yet even then, I think he followed the same line of thinking from my interpretation to make this story up. Or in other words, even if the story is completely invented, I think the journalist created it BECAUSE of the github leak rather than despite it. Of course, that proves nothing, the only thing I was defending is that both rumors fit well together, not the veracity of either.

 

[Kaotik]

1 - 33% difference is gigantic. The PS4 Pro used more advanced (and noticeably bigger) CUs and had dedicated hardware for upscaling, yet the difference between the XBoneX and PS4Pro is still big.

Care to elaborate on this? I mean, every single GPU since forever offers "hardware upscaling" and MS utilizes this on Xbox Ones too (not (just) X)

 

[Shifty Geezer]

With articles like this, I don't take any specific word or number at face value, but rather I keep the general narrative as a possibility in the back of my head.

And that contrasts with technical articles like a DF interview where particular words and phrases may be scrutinised more carefully (if not coming from PR mouthpieces). Hence the nit-picking about whether PS4 was an SSD or not after the Wired article, where it mattered who the term 'SSD' originated from; as Architect of PS5, Cerny's words are gospel, but one has to be particular about what words are his and what are the reporter's to know which words are gospel and which apocryphal, so to speak.

 

[Shifty Geezer]

Care to elaborate on this? I mean, every single GPU since forever offers "hardware upscaling" and MS utilizes this on Xbox Ones too (not (just) X)

I guess he means the ID buffer intended to facilitate reconstruction techniques. 'Hardware for upscaling' doesn't equate to 'upscaling hardware'.

 

[anexanhume]

I guess he means the ID buffer intended to facilitate reconstruction techniques. 'Hardware for upscaling' doesn't equate to 'upscaling hardware'.

Speaking of ID buffers

Index buffer block compression - Microsoft Technology Licensing

http://www.freepatentsonline.com/y2020/0051286.html

 

[disco_]

You're assuming the surprise aspect of the BoM amounts to no less than $80 over what they originally intended (which would be the same $370 as the OG PS4). That's a lot for a surprise.

What components would you believe was expected to cost X, but ended up being X+$80?
The article mentions NAND and RAM, but a quick visit to dramexchange will tell you there hasn't been any price hike on either front, on the contrary.

Could RAM and NAND still be more expensive than Sony predicted? Yes, but I doubt they'd miss the mark as far as $80.

I'll concede that maybe Sony was aiming for a $420-430 BoM so they could sell for $450 with almost no margins at first (like the PS4), and a $450 BoM puts them in a tough spot deciding whether to go up to $499 with clear profit or $449 with losses.

But there's nothing in that $450 BoM that says Sony was aiming at specifically $399, on the contrary.

If we look at bloomberg, it's pretty obvious sony decided to forego the usual <$1 cooler and splurged to put in a noctua nh-d15. With their discount they could easily get that down and boom, there's your $80 increase.

 

[AbsoluteBeginner]

It doesn't proof anything, it's only doing regression tests for backwards compatibility with PS4 and PS4 Pro. That one line about being "full chip" might even mean something different to who wrote it, other than full PS5 chip. It's your interpretation, not proof of anything.

And what would be the point of going to the trouble of developing a 36CU chip to test it at 2Ghz if another one with more CUs and lower clock would be coming down the line still targeting 9TF? In your effort to insist with 9TF you start not making much sense.

No, its regression test for rumored PS5 chip using BC1, BC2 and Native modes, therefore its not BC regression test but regression test of GPU in all 3 modes of work (PS4 BC, Pro BC and Native).

Again, patently false arguing that Github shows nothing. In fact, Arden also has Native, BC1 and BC2 modes and the way people got 56CU number is by using Native mode that shows it.

 

[AbsoluteBeginner]

If we look at bloomberg, it's pretty obvious sony decided to forego the usual <$1 cooler and splurged to put in a noctua nh-d15. With their discount they could easily get that down and boom, there's your $80 increase.

People argue Sony can and should recoup deficit of 600$ console that costs 500$ by PS Plus and online, but I think Sony had different plan in their mind.

I think they wanted to recoup deficit of console with ~$420 BOM while selling it for $399. Same strat as PS4 one, but selling it at a higher loss yet still maintaing fantastic price/performance ratio. Unbeatable one actually. It would have BC with PS4/Pro, it would be extremely good deal for a money and while they would lose a bit on each console sold, they would have recoup it back pretty much instantly while making sure most customers get on board ASAP.

What could have happened, if we assume Bloomberg report is correct, is that original BOM went a higher then expected and put them in weird spot.

If they sell for $399 they lose too much per console, if they sell it for $499 they lose price/perf advantage they would have held at $399. That way they dont lose money, but go head to head with competitor that always shot at higher BOM and price, thus higher specs.

If we look into rumored PS5, apart from chip itself, its all therebouts with XSX. So they really could have gone with smaller and cheaper chip but matching XSX everywhere else. If MS went with 20GB of GDDR6, I doubt 16GB of GDDR6 (highest speed modules at that) with 4GB of DDR4 would be considerably cheaper option.

If they wanted to save money, they must have looked at chip itself and figured out that at 7nm they save save substantial amount of money by going with 320-330mm² chip instead of 400mm² one.

We dont know, but IMO everything points to that direction.

I always said $399 PS5, even with 9.2TF Navi, is very tall order and that by selling it at that price Sony would position them in prefect spot. Think PS4 Pro vs XBX only without pretty much every advantage Scorpio has over Pro and few actual advantages PS5 would hold while being $100 dollars cheaper.

Therefore my theory is :

Sony aimed at ~$420 BOM (compared to $380 for PS4) while selling it at amazing price of $399, thus loosing ~60-70$ per console.

If price spikes, which Sony DID allude to, have happened, then they will be looking additional deficit on every console sold and that is not a great prospect for shareholders since early strat with lower BOM was still very agressive. Thus, Sony will be waiting on MS to see how they should price their console.

MS probably always aimed for $499 and higher BOM, but NOT much higher where they would be hurt on price if spike was to happen. Basically, XSX would cost 100$ more while original BOM was ~40$ more, thus even with spike their strat was not as price agressive for them to care.

 

[disco_]

People argue Sony can and should recoup deficit of 600$ console that costs 500$ by PS Plus and online, but I think Sony had different plan in their mind.

I really don't buy the huge loss strategy from either side.

 

[Picao84]

People argue Sony can and should recoup deficit of 600$ console that costs 500$ by PS Plus and online, but I think Sony had different plan in their mind.

I think they wanted to recoup deficit of console with ~$420 BOM while selling it for $399. Same strat as PS4 one, but selling it at a higher loss yet still maintaing fantastic price/performance ratio. Unbeatable one actually. It would have BC with PS4/Pro, it would be extremely good deal for a money and while they would lose a bit on each console sold, they would have recoup it back pretty much instantly while making sure most customers get on board ASAP.

What could have happened, if we assume Bloomberg report is correct, is that original BOM went a higher then expected and put them in weird spot.

If they sell for $399 they lose too much per console, if they sell it for $499 they lose price/perf advantage they would have held at $399. That way they dont lose money, but go head to head with competitor that always shot at higher BOM and price, thus higher specs.

If we look into rumored PS5, apart from chip itself, its all therebouts with XSX. So they really could have gone with smaller and cheaper chip but matching XSX everywhere else. If MS went with 20GB of GDDR6, I doubt 16GB of GDDR6 (highest speed modules at that) with 4GB of DDR4 would be considerably cheaper option.

If they wanted to save money, they must have looked at chip itself and figured out that at 7nm they save save substantial amount of money by going with 320-330mm² chip instead of 400mm² one.

We dont know, but IMO everything points to that direction.

I always said $399 PS5, even with 9.2TF Navi, is very tall order and that by selling it at that price Sony would position them in prefect spot. Think PS4 Pro vs XBX only without pretty much every advantage Scorpio has over Pro and few actual advantages PS5 would hold while being $100 dollars cheaper.

Therefore my theory is :

Sony aimed at ~$420 BOM (compared to $380 for PS4) while selling it at amazing price of $399, thus loosing ~60-70$ per console.

If price spikes, which Sony DID allude to, have happened, then they will be looking additional deficit on every console sold and that is not a great prospect for shareholders since early strat with lower BOM was still very agressive. Thus, Sony will be waiting on MS to see how they should price their console.

MS probably always aimed for $499 and higher BOM, but NOT much higher where they would be hurt on price if spike was to happen. Basically, XSX would cost 100$ more while original BOM was ~40$ more, thus even with spike their strat was not as price agressive for them to care.

I don't have any difficulty in thinking about a 9.2TF console selling for 399. It makes sense. What does not make sense, at all, is planning for a GPU clocked at 2 GHz. Especially after very conservative clocks on PS4 and PS4 Pro. Especially with completely new GPU tech like Ray Tracing and other stuff. It's a massive gamble on Sony's part, who were burned in the not too long distant past with PS3 and Cell's woes.

Regarding GitHub, I didn't say that it shows nothing, I said it doesn't proof anything. They are not one and the same. GitHub leak shows what it shows, it doesn't have to mean that is telling us the full truth. Yes, it does mention a Gen 2 (does it actually have the word native on it at all??), but we can only assume that that is PS5, we don't really know for sure. It might even be a chip for PS Now for all we know or for some exotic product they only sell in Japan. Plus the fact that this chip reads as Navi 10 Lite and Navi 10 does not have ray tracing at all! That should be a big red herring that something is amiss.

 

[iroboto]

Plus the fact that this chip reads as Navi 10 Lite and Navi 10 does not have ray tracing at all! That should be a big red herring that something is amiss.

As we saw with 4Pro which is distinctly Polaris. They back ported features from
Vega. Like Rapid Packed Math.

So I believe this should be doable for Sony to do again if they choose to.

I don’t necessarily believe that generations of architecture may matter as much in the console space as they do in the PC space. Just given the nature of a semi custom chip. being able to pick and choose and make adjustments for your needs seems more important than moving your architecture up another digit.

 

[AbsoluteBeginner]

I really don't buy the huge loss strategy from either side.

I agree, I dont buy it because business in 2020 is much different from 2002 or 2006 for that matter. Digital, streaming, mid refresh, play everywhere etc. essentially put an end to arms race with serious price deficit. Its all about getting as much users as possible on your platform and keep it that way. Slow ramp up, loss strat, it makes no sense in todays world.

 

[Picao84]

As we saw with 4Pro which is distinctly Polaris. They back ported features from
Vega. Like Rapid Packed Math.

So I believe this should be doable for Sony to do again if they choose to.

I don’t necessarily believe that generations of architecture may matter as much in the console space as they do in the PC space. Just given the nature of a semi custom chip. being able to pick and choose and make adjustments for your needs seems more important than moving your architecture up another digit.

Didn't someone show a few pages ago that Navi10 Lite was showing up on MacOS as well? Therefore it's not a Sony exclusive SKU, it doesn't point to a custom part for Sony.

 

[temesgen]

And that contrasts with technical articles like a DF interview where particular words and phrases may be scrutinised more carefully (if not coming from PR mouthpieces). Hence the nit-picking about whether PS4 was an SSD or not after the Wired article, where it mattered who the term 'SSD' originated from; as Architect of PS5, Cerny's words are gospel, but one has to be particular about what words are his and what are the reporter's to know which words are gospel and which apocryphal, so to speak.

The Gospel of Klee is considered a non canonical book yet my research into apocryphal text shows a surprising amount of support for the GitHub scrolls, a book or pages that do have carbon dating to support their age but are incomplete and lack context to derive their complete meaning

 

[Picao84]

Tin foil hat moment:
What if Oberon is actually a desktop/laptop part for some exotic product Sony is launching? A PC that can play PS4 games as well as PC games? Some odd answer to Microsoft strategy of supporting the Xbox ecosystem in both consoles and PCs.

(Don't take this a serious though, I know chances of this are 0,01%)