• Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.

AegonSnake

Banned
Oct 25, 2017
9,566
Although your post is kinda hyperbolic and I don't agree with your price conclusions because I think Sony is willing to eat the extra cost. However, as someone who has been closely paying attention to the tech side of these consoles this post has a lot of truth to it. Sony going with a 2.23 clock was completely shocking. To support BC, Sony went fast and narrow and, from a business perspective, it's looking to be the wrong gamble. I think when we look back on these consoles in like 5 years, the PS5 technology design will be considered something of a disappointment. Not a disappointment for most gamers, but a disappointment for people who analyze this stuff from a tech and business perspective. It isn't all bad news - the one possible upside for the PS5 is long term they may be able to more aggressively reduce costs than the XSX due to the lower CU count, etc.

That said, I think Sony is willing to eat these costs now to sell a ton of consoles, and for most gamers none of this will really matter. However, I think Sony biz folks and shareholders aren't loving the PS5 design right now and I suspect Sony is regretting some of their decisions.

IMO, of course.
Well, like i said we will find out tomorrow and its possible Sony will eat up the costs. People dont realize this but while the PS4 launch at $399, we still had to pay an extra $60 a year just to play online. that was a new cost that was passed down to consumers. if we are expecting Sony to take a $450 BOM console with 50% to still unstable yields and launch it at $400, we should also expect to pay for some of those losses.
 

DrDeckard

Banned
Oct 25, 2017
8,109
UK
If sony get to $399 it's absolutely insane. that thing will fly. It will be pandemonium!

I hope we see it, the crazy buggers!
 

Nameless

Member
Oct 25, 2017
15,347
Hot PS5 devkits dont indicate poor SoC yields?

I am failing to see the link to reliability here. I made no such comments.

I'm asking how you get to a place where you now believe ALL of the FUD around PS5? Which did include unsubstantiated reports of Sony panicking over failing units.

And dev kits can run hot for a number of reasons. The Pro runs hot and we never heard about yield issue.
 

Alexandros

Member
Oct 26, 2017
17,800
lmao. I am used to it by now. My posts cut right through people. It's a talent. or a curse.... I cant tell.

I hope they tag me tomorrow, i dont want this console to be $599, but right now with what we know, i cannot reconcile a $399 console in my mind unless sony is willing to take a massive loss or pass down the costs to the consumers somehow. The BOM was already $70 higher than the PS4 BOM so at $399 they would already be taking a $50-70 extra loss. Regardless, The estimates are down by 4 million units, failure rate was as low as 50% and still not stable. if this isnt the time to panic, i dont know what is. You were in the next gen threads with us, everyone dismissed the github leaks because even 2.0 ghz was too high and we all knew the yields would be bad. This is all our worst fears come true.

I wish I had the same faith in Sony like these guys do, but this is arguably the worst news we couldve heard going into the conference tomorrow.

but-why.gif


I don't see any issue. If something went wrong in the manufacturing process and yields were not as expected then Sony will eat the cost. There's no way it will increase the console's price and give Microsoft a price advantage on both models.
 
Oct 25, 2017
1,234
UK NW
Either way they still have a lot of consoles to sell, demand will far exceed the amount produced and I'd expect huge shortages after 3-6 months.
That GPU upclock to 2.23GHZ is easily the most obvious cause of the lower than expected yields though.
Most AMD GPU's settle around less than 2GHZ chips unless they are pre-binned chips.

If Sony does eat the cost for the lower than expected yields plus the loss for building the consoles as a whole then they must expect yields to be become much better/attainable by the end of next year or sooner.
 

Jade1962

Banned
Oct 28, 2017
4,259
Do American reporters not look into console production. Seems like we ever only hear stories about Nintendo and Sony manufacturing. 11 million by March seems like a lot to me but maybe MS will be supplying more Xboxes than Sony PS5s.
 

platocplx

2020 Member Elect
Member
Oct 30, 2017
36,072
Do American reporters not look into console production. Seems like we ever only hear stories about Nintendo and Sony manufacturing. 11 million by March seems like a lot to me but maybe MS will be supplying more Xboxes than Sony PS5s.
Yeah Apple has had some issues with the iPhone and there have been months of rumors no announcement till October because of it. Nothing huge but even them having their own issues just says there are some manufacturing impacts
 
Oct 25, 2017
4,156
Well, like i said we will find out tomorrow and its possible Sony will eat up the costs. People dont realize this but while the PS4 launch at $399, we still had to pay an extra $60 a year just to play online. that was a new cost that was passed down to consumers. if we are expecting Sony to take a $450 BOM console with 50% to still unstable yields and launch it at $400, we should also expect to pay for some of those losses.

You might be right. Sony is cutting the supply which means they can charge higher prices. I wouldn't be shocked if they went with 499/549, but I think the prices will be lower. Recently, there's a lot of insider smoke around the 399 price point. I just think Sony is willing to take a bath on these consoles while they work reduce their costs.

Like I said, I seriously doubt Sony is loving the PS5 design right now. It's not going to print as much money as the PS4 on launch. The PS4 was a legit cost marvel when it debuted. A slightly above cost at launch is a console manufactures dream. The PS5 isn't shaping up to be that...at all.
 
Last edited:

Fezan

Member
Oct 26, 2017
3,274
Ok just talked to a friend who knows about manufacturing. He had some positive and neagitive thoughts about it

- 50% yeilds are on lower side but not unheard of in high end chips. He estimates Xbox one chip had 65% yeilds in the beginning even with smaller number of Cu and much more mature design likely due to esram.

- moving higher than 50% yeilds means that they are on the right path and have found a solution for it.

- yeilds are not linear. They are mostly curve where yeilds improve a lot more with passage of time. In a year they will most probably each industry standard .

- 50% yeilds doesn't mean half of all the wafers are wasted. It just means that batch of wafer had this issue. Later batches can have much higher yields.

- lower yeilds are less dependent on clocks as these are the first things tested in intial run and more with complexity / voltagrof the chips they are making.

- lastly if Sony has decreased the number of orders it means Sony is waiting for yeilds to improve hence facing smaller loss.

Hope this answers some of the questions here. I can ask him more of you guys want to know.
 
Last edited:

LordBlodgett

Member
Jan 10, 2020
806
So the narrative now is "high clocks = bad yields". Now everything makes sense. That is why I paid $70 less for my custom RTX2080 that clocks 300MHz higher compared to the FE model out of the box. Man MSI must've lost a lot of money for every GPU they sold with their bad yields because of the higher clocks... /s

High clocks do not equal lower yields. If a chip is designed to reach a specific clock within a thermal budget, all chips that are not damaged on a wafer will reach that clock within that budget. Some will be better (=cooler), some will be worse (=hotter), but all will work within spec. You can switch thermal budget for clock budget and fixed thermals and you get the same but with differing clock speeds.
This is absolutely wrong. For either graphics cards or CPUs you will have chips that meet spec or don't. In most of these cases the highest quality chips get put in top of the line parts, and low speed chips get put in the low end parts. This is called "binning" and is perfectly normal for all chips. The difference with consoles is that they don't have a lower end model to bin chips to
 
Oct 27, 2017
4,919
This is absolutely wrong. For either graphics cards or CPUs you will have chips that meet spec or don't. In most of these cases the highest quality chips get put in top of the line parts, and low speed chips get put in the low end parts. This is called "binning" and is perfectly normal for all chips. The difference with consoles is that they don't have a lower end model to bin chips to

So essentially, any functional chips that don't meet the clock/voltage/etc target would not be listed as a yield problem of the wafer but would still get thrown away?
 

Deleted member 35478

User-requested account closure
Banned
Dec 6, 2017
1,788
Microsoft are still desperate to have price mic drop, so wouldn't surprise me if they reduce the Series X to $459 if the PS5 SE was to come in at $499. The problem of course is that it wouldn't make any difference if the DE is $399

I don't see MS lowering the Xbox prices in reaction to the PS5 pricing, but I do see them bundling GamePass (6-12 months) with each Xbox to counter PS5 prices if necessary.
 

Chamon

Member
Feb 26, 2019
1,221
Ok just talked to a friend who knows about manufacturing. He had some positive and neagitive thoughts about it

- 50% yeilds are on lower side but not unheard of in high end chips. He estimates Xbox one chip had 65% yeilds in the beginning even with smaller and much more mature design.

- moving higher than 50% yeilds means that they are on the right path and have found a solution for it.

- yeilds are not linear. They are mostly curve where yeilds improve a lot more with passage of time. In a year they will most probably each industry standard .

- 50% yeilds doesn't mean half of the wafer is wasted. It just means that batch of wafer had this issue. Later batches can have much higher yields.

- lower yeilds are less dependent on clocks as these are the first things tested in intial run and more with complexity of the chips they are making.

- lastly if Sony has decreased the number of orders it means Sony is waiting for yeilds to improve hence facing smaller loss.

Hope this answers some of the questions here. I can ask him more of you guys want to know.
Thank you. This is very informative.
 

MrFox

VFX Rendering Pipeline Developer
Verified
Jun 8, 2020
1,435
Ok just talked to a friend who knows about manufacturing. He had some positive and neagitive thoughts about it

- 50% yeilds are on lower side but not unheard of in high end chips. He estimates Xbox one chip had 65% yeilds in the beginning even with smaller and much more mature design.

- moving higher than 50% yeilds means that they are on the right path and have found a solution for it.

- yeilds are not linear. They are mostly curve where yeilds improve a lot more with passage of time. In a year they will most probably each industry standard .

- 50% yeilds doesn't mean half of the wafer is wasted. It just means that batch of wafer had this issue. Later batches can have much higher yields.

- lower yeilds are less dependent on clocks as these are the first things tested in intial run and more with complexity of the chips they are making.

- lastly if Sony has decreased the number of orders it means Sony is waiting for yeilds to improve hence facing smaller loss.

Hope this answers some of the questions here. I can ask him more of you guys want to know.
Thanks for the insights, it makes a lot of sense.

So if XB1 at 28nm was 65% yield, and if 7nm is supposed to be lower yield (from microsoft presentation), it's easier to accept 50% being a reasonable figure.

Also, it all adds up to confirm what we heard about the difficulties at 7nm, the EUV delays made the situation less than ideal.
semiengineering.com

7nm Fab Challenges

7nm Fab Challenges FinFET formation, mask challenges and back-end-of-line issues will make this node difficult and expensive.
 
Oct 25, 2017
4,156
I remember people here - including me - laughing at him and GitHub... That didn't age well lol

Yeah, I think Blue Nugroho is passionate and perhaps a bit hyperbolic, which easy to confuse with being a fanboy. I was a skeptic just like you, but history has been kind to him.

I'm also starting to wonder if his hypothesis that the PS5 doesn't use the same RNDA 2 like the XSX is also true.



I'm looking forward to the PS5 teardown.
 

jroc74

Member
Oct 27, 2017
28,992
I remember people here - including me - laughing at him and GitHub... That didn't age well lol
Yeah, I think Blue Nugroho is passionate and perhaps a bit hyperbolic, which easy to confuse with being a fanboy. I was a skeptic just like you, but history has been kind to him.

I'm also starting to wonder if his hypothesis that the PS5 doesn't use the same RNDA 2 like the XSX is also true.



I'm looking forward to the PS5 teardown.


Oh wait...you're serious....?

Eh...
 

sjackso3

Member
Oct 30, 2017
630
Houston
Yeah, I think Blue Nugroho is passionate and perhaps a bit hyperbolic, which easy to confuse with being a fanboy. I was a skeptic just like you, but history has been kind to him.

I'm also starting to wonder if his hypothesis that the PS5 doesn't use the same RNDA 2 like the XSX is also true.



I'm looking forward to the PS5 teardown.


I'm not not as tech savvy as some on here. What does this mean?
 

Praedyth

Member
Feb 25, 2020
6,519
Brazil
Yeah, I think Blue Nugroho is passionate and perhaps a bit hyperbolic, which easy to confuse with being a fanboy. I was a skeptic just like you, but history has been kind to him.

I'm also starting to wonder if his hypothesis that the PS5 doesn't use the same RNDA 2 like the XSX is also true.



I'm looking forward to the PS5 teardown.

Did AMD produce two RDNA 2? What?
 
Oct 25, 2017
4,156
Oh wait...you're serious....?

Eh...

I'm not saying it's true, but it could help explain the low yields.

Also, I'm seeing folks outright dismiss Blue Nugroho theories, but very little to backup why he's wrong. Just a lot of "LOL" crap. He's been right about a lot of stuff. His analysis deserves way more respect and attention.

But again, I have no idea if it's true or not.
 

sjackso3

Member
Oct 30, 2017
630
Houston
If I understand correctly that their are two versions of RDNA 2 (which seems wierd), and the PS5 is based on a weaker version than the Xbox Series line, then maybe the PS5 will come in cheaper than we think. That would be a win for Sony IMO. At the end of the day, if the games look and play good, nobody will care.
 

Mubrik_

Member
Dec 7, 2017
2,723
The top link literally says that 80 percent as a baseline to 90 percent plus is needed. 50 percent is atrociously low yields on an established node in that case.
Yeah but for a mature product
These are pretty new.

lmao. I am used to it by now. My posts cut right through people. It's a talent. or a curse.... I cant tell.

I hope they tag me tomorrow, i dont want this console to be $599, but right now with what we know, i cannot reconcile a $399 console in my mind unless sony is willing to take a massive loss or pass down the costs to the consumers somehow. The BOM was already $70 higher than the PS4 BOM so at $399 they would already be taking a $50-70 extra loss. Regardless, The estimates are down by 4 million units, failure rate was as low as 50% and still not stable. if this isnt the time to panic, i dont know what is. You were in the next gen threads with us, everyone dismissed the github leaks because even 2.0 ghz was too high and we all knew the yields would be bad. This is all our worst fears come true.

I wish I had the same faith in Sony like these guys do, but this is arguably the worst news we couldve heard going into the conference tomorrow.
reserved you a section


The best reason to have some shipped by air is because the production is continuous, so they ship the first batches by boat, and they can also ship the last weeks of production before the launch date by air, which ends up the maximum possible consoles at launch for the lower possible cost.

I think they always did that with every console launch?
Yup, apparently they did for PS4 and Nintendo did as well.
 

rustyra24

Member
Jul 6, 2018
468
I am wondering how many chips that passed QC could be on the edge? This makes me worry about reliability. i will probably get it anyways.
 

Dphex

Member
Oct 27, 2017
12,811
Cologne, Germany
The big thing many people are missing is that Sony does not need to match Xbox pricing in any way, they can go more expensive and the console will still sell a lot.
 

Mubrik_

Member
Dec 7, 2017
2,723
The big thing many people are missing is that Sony does not need to match Xbox pricing in any way, they can go more expensive and the console will still sell a lot.
Hear this alot, even though you are correct, most gamers are sensitive to pricing.

They lost a ton of market share with the PS3 majorly due to delay and pricing,
Keep in mind that the hardware sales is not where they make most profits from but the ecosystem, even if they can go more expensive they most likely would not because they might/would be slowing down sales/adoption that would lead to larger profits from player spending.

And the gamers will say "But not as much as the new Xbox!"
Don't think sony cares about that bit save for it makes for good PR lol
 

Sia

Attempted to circumvent ban with alt account
Banned
Jun 9, 2020
825
Canada
Yeah, I think Blue Nugroho is passionate and perhaps a bit hyperbolic, which easy to confuse with being a fanboy. I was a skeptic just like you, but history has been kind to him.

I'm also starting to wonder if his hypothesis that the PS5 doesn't use the same RNDA 2 like the XSX is also true.



I'm looking forward to the PS5 teardown.

There was that italian sony engineer that said PS5 is missing 1 RDNA2 feature but has several other custom features added in.
 

platocplx

2020 Member Elect
Member
Oct 30, 2017
36,072
Ok just talked to a friend who knows about manufacturing. He had some positive and neagitive thoughts about it

- 50% yeilds are on lower side but not unheard of in high end chips. He estimates Xbox one chip had 65% yeilds in the beginning even with smaller and much more mature design.

- moving higher than 50% yeilds means that they are on the right path and have found a solution for it.

- yeilds are not linear. They are mostly curve where yeilds improve a lot more with passage of time. In a year they will most probably each industry standard .

- 50% yeilds doesn't mean half of the wafer is wasted. It just means that batch of wafer had this issue. Later batches can have much higher yields.

- lower yeilds are less dependent on clocks as these are the first things tested in intial run and more with complexity of the chips they are making.

- lastly if Sony has decreased the number of orders it means Sony is waiting for yeilds to improve hence facing smaller loss.

Hope this answers some of the questions here. I can ask him more of you guys want to know.
this sounds pretty reasonable and fair.
 

B.O.O.M.

Member
Oct 25, 2017
8,757
Ok just talked to a friend who knows about manufacturing. He had some positive and neagitive thoughts about it

- 50% yeilds are on lower side but not unheard of in high end chips. He estimates Xbox one chip had 65% yeilds in the beginning even with smaller and much more mature design.

- moving higher than 50% yeilds means that they are on the right path and have found a solution for it.

- yeilds are not linear. They are mostly curve where yeilds improve a lot more with passage of time. In a year they will most probably each industry standard .

- 50% yeilds doesn't mean half of the wafer is wasted. It just means that batch of wafer had this issue. Later batches can have much higher yields.

- lower yeilds are less dependent on clocks as these are the first things tested in intial run and more with complexity of the chips they are making.

- lastly if Sony has decreased the number of orders it means Sony is waiting for yeilds to improve hence facing smaller loss.

Hope this answers some of the questions here. I can ask him more of you guys want to know.

Thank you. This was really informative
 

Brohan

The Fallen
Oct 26, 2017
2,544
Netherlands
I'm not saying it's true, but it could help explain the low yields.

Also, I'm seeing folks outright dismiss Blue Nugroho theories, but very little to backup why he's wrong. Just a lot of "LOL" crap. He's been right about a lot of stuff. His analysis deserves way more respect and attention.

But again, I have no idea if it's true or not.

Eh that person is getting laughed at for a reason. It's Misterxmedia level bullshit. Actually I am convinced that it's most likely the same person.
 

Mubrik_

Member
Dec 7, 2017
2,723
Im honored but jroc had me in three sections.

2CXD0v.png
And he's right on all 3

Ok just talked to a friend who knows about manufacturing. He had some positive and neagitive thoughts about it

- 50% yeilds are on lower side but not unheard of in high end chips. He estimates Xbox one chip had 65% yeilds in the beginning even with smaller and much more mature design.

- moving higher than 50% yeilds means that they are on the right path and have found a solution for it.

- yeilds are not linear. They are mostly curve where yeilds improve a lot more with passage of time. In a year they will most probably each industry standard .

- 50% yeilds doesn't mean half of the wafer is wasted. It just means that batch of wafer had this issue. Later batches can have much higher yields.

- lower yeilds are less dependent on clocks as these are the first things tested in intial run and more with complexity of the chips they are making.

- lastly if Sony has decreased the number of orders it means Sony is waiting for yeilds to improve hence facing smaller loss.

Hope this answers some of the questions here. I can ask him more of you guys want to know.
Sounds solid.
If anything it will be about the relationship between yield and the APU cost
Personally, i think sony factored this in already but someone knowledgeable should sway all the FUD or better yet we wait for tomorrow lol

Some info on TSMC and semiconductor design, production, and delivery, as it seems there is VAST misinformation here.

1- A third party such as Nvidia, AMD, etc, put forth a design, which is then heavily cooperatively developed for the process node that is available for the delivery of final product. This process often requires a number of 'respins', where errata and optimization efforts refine the final product.

2- Upon reaching final or near-final, samples are cut and sent to unit manufacturing lines so they can place in PCB, validate thermals, power delivery, firmware, bus and memory interface design and behavior, etc. Ideally when this goes well, the first bulk orders are placed with TSMC, going from a reservation to a negotiated price for delivery of X products per contract. Yields are ALREADY in the equation at this point, and are the responsibility of TSMC to deliver based on the preliminary qualification sample run and 'risk' production statistics.

3- TSMC runs the fabrication processes for bulk, in this case 300mm 7nm, resulting in X number of dies per wafer before final yields reduce to available dies that pass. It is at TSMC during the production phase that dies are tested, cut, etched, and depending on the contract, packaged accordingly (from nearly bare, as in the case with AMD's Ryzen chiplets, to availability of nearly complete BGA packaged or capped processors ready to mount).

In the case of Sony and Microsoft, they are receiving working individual SoCs, already cut, etched, and validated by TSMC. It would be horrifically nonsensical and inefficient to ship 300mm wafers to Sony and MS manufacturing partners just to have to be tested, cut, and packaged after the fact. That's absurdly stupid to even consider, especially given TSMC's expertise in dealing with their own leading edge node production and engineering end-to-end. It would be like shipping random boxes of untested, incomplete transmissions from supplier to BMW or Ford rather than final and validated ones ready to install, and also cause chaos and lawsuits related to contract and cost.
This post sheds some light on that as well.
 

19thCenturyFox

Prophet of Regret
Member
Oct 29, 2017
4,309
Whenever I lowball predictions or when I'm perhaps a bit overly skeptic I tell myself "Man, if you're wrong - awesome - and if you're right, shit sucks, but you at least get to tell everyone you were right and derive some enjoyment from that". Well I was right and that doesn't make it any better. Might have to think this strategy over.
 
Oct 25, 2017
4,156
Eh that person is getting laughed at for a reason. It's Misterxmedia level bullshit. Actually I am convinced that it's most likely the same person.

He's no Misterxmedia. I honestly can't recall him being wrong about anything. I think the only thing that surprised everyone, including him, were the PS5 clock speeds. I don't think anyone thought Sony would attempt that. These yields might be a good reason why.
 

Brohan

The Fallen
Oct 26, 2017
2,544
Netherlands
He's no Misterxmedia. I honestly can't recall him being wrong about anything. I think the only thing that surprised everyone was the PS5 clock speeds. I don't think anyone thought Sony would attempt that. These yields might be the reason why.

The dude is wrong about a ton of stuff all the time. He's still trying to claim that the PS5 isn't actually RDNA2. The dude is either Misterxmedia or Misterxmedia's long lost brother or something.
 

Shogmaster

Banned
Dec 12, 2017
2,598
Yeah, I think Blue Nugroho is passionate and perhaps a bit hyperbolic, which easy to confuse with being a fanboy. I was a skeptic just like you, but history has been kind to him.

I'm also starting to wonder if his hypothesis that the PS5 doesn't use the same RNDA 2 like the XSX is also true.



I'm looking forward to the PS5 teardown.

LOL who's gonna do a teardown with an X ray microscope to see how the PS5 SoC works. We can only take Sony and AMD's word.

I;m more looking forward to XSS teardown and see how the 10GB of RAM is actually laid out.
 

MrFox

VFX Rendering Pipeline Developer
Verified
Jun 8, 2020
1,435
Remember, if a leak ends up completely wrong, or contradict the previous leak, it's because plans changed. All leaks are always true.
 
Oct 25, 2017
4,156
The dude is wrong about a ton of stuff all the time. He's still trying to claim that the PS5 isn't actually RDNA2. The dude is either Misterxmedia or Misterxmedia's long lost brother or something.

The PS5 is RNDA 2 because that's how AMD is presenting it. So the PS5 absolutely has it. He's saying the PS5's RNDA 2 version isn't the same as the newer XSX/XSS version. He believes the PS5 version is partly based on RNDA 1.x tech.

Also, some "this is just absurd" posts haven't aged well. Remember when Sony going over 2.x for the clock speed was crazy talk and that means the PS5 MUST have more CUs?

I don't know what is true, but I certainly have an open mind about it. I'm really looking forward to more PS5 details.
 

Brohan

The Fallen
Oct 26, 2017
2,544
Netherlands
The PS5 is RNDA 2 because that's how AMD is presenting it. So the PS5 absolutely has it. He's saying the PS5's RNDA 2 version isn't the same as the newer XSX/XSS version. He believes the PS5 version is partly based on RNDA 1.x tech.

Also, some "this is just absurd posts" haven't aged well. Remember when Sony going over 2.x for the clock speed was crazy talk and that means the PS5 MUST have more CUs?

I don't know what is true, but I certainly have an open mind about it. I'm really looking forward to more PS5 details.

Either the PS5 is RDNA2 or it isn't. There is no RDNA 1.5 which is why his tweets are ridiculous.

Manufacturers may instead opt to leave out features they don't deem necessary or instead add ones that aren't actually part of the standard specifications. These are completely customised chips.
It may very well be a possibility that because of said custimizations there could be differences in feature sets between MS and Sony.

These custimizations do not however make it RDNA "1.5" or RDNA3.
 

DukeBlueBall

Banned
Oct 27, 2017
9,059
Seattle, WA
Also, I'm seeing folks outright dismiss Blue Nugroho theories, but very little to backup why he's wrong. Just a lot of "LOL" crap. He's been right about a lot of stuff. His analysis deserves way more respect and attention.

He has been right about stuff that was already right.
He takes the correct info and then go conclusion hunting.

He believed PS5 has only software RT and used Github/etc to prove it when we all knew not all the info is complete.

He estimates Xbox one chip had 65% yeilds in the beginning even with smaller and much more mature design.

This part is actually not correct. Xb1 SOC is 370mm+. It is bigger than Xsx (~360mm) and much bigger than PS5's SOC (~300m).