Support NeoGAF

[M1chl]

Its funny,I just read through all pages and its you xbox guys starting the drama, then you guys cry wolf. Kinda sad, really

What "you xbox guys", I just trying to moderate my own thread. Even those "xbox guys" should shut the fuck up about PS5 in this thread and don't bring XSX to the potentiall PS5 hot chip thread.

'Cause otherwise, this would be just another Next-gen speculation thread.

Imo we need more credible feedback (less rumors) on the matter to even think something.

We mainly need input from dev. From my perspective, it was harder to make performance sensitive iOS app then Android app (this was harder to test obviously), because dev tools were better on Android. So from my perspective, it depends heavily on dev tools. Which we don't heard anything so far. Is the documentation is going to be available for everyone? Etc. There are way more variables to the development than just 2 HW targets vs 1. Obviously Sony can be better at both and then MS will suffer. Just saying my perspective.

 

[psorcerer]

Well, Phil actually play games. And I do think that it's vastly overrated how much potential XSS will compicate development.

It is not.
Supporting each new platform is a nightmare.
Making games stand out on new hardware while supporting old one is nightmare^2

 

[Dodkrake]

For anybody that has a problem grasping why more platforms to be supported is a shitshow in development, look no further than iOS vs Android. Android needs to cover a wide range of devices, iOS is way more limited. This allows Apple to have full control of the Hardware and optimize the shit out of it, getting away with things such as Less Ram, lesser displays, etc, and still have a similar and sometimes better outcome.

 

[Jon Neu]

Its funny,I just read through all pages and its you xbox guys starting the drama, then you guys cry wolf. Kinda sad, really

It's funny; you see the first page of the thread and is amazing. A positive thread about XsX in GAF and is not being trolled or dismissed? I almost cried, but then I thought: how much it would last?

Then you see the next pages with the introduction of the usual Sony posters and everything changes.

It really makes you think.

 

[M1chl]

It is not.
Supporting each new platform is a nightmare.
Making games stand out on new hardware while supporting old one is nightmare^2

Except this time is basically watered down version of the same HW, with same feature set, etc.

For anybody that has a problem grasping why more platforms to be supported is a shitshow in development, look no further than iOS vs Android. Android needs to cover a wide range of devices, iOS is way more limited. This allows Apple to have full control of the Hardware and optimize the shit out of it, getting away with things such as Less Ram, lesser displays, etc, and still have a similar and sometimes better outcome.

We mainly need input from dev. From my perspective, it was harder to make performance sensitive iOS app then Android app (this was harder to test obviously), because dev tools were better on Android. So from my perspective, it depends heavily on dev tools. Which we don't heard anything so far. Is the documentation is going to be available for everyone? Etc. There are way more variables to the development than just 2 HW targets vs 1. Obviously Sony can be better at both and then MS will suffer. Just saying my perspective.

Hmm, that's not it. One think which I know for sure, that displays on Android is non-issue and iOS layouts are nightmare even at few displays they have.

 

[jimbojim]

Posted previous page back, what we know so far....points as so...

Sony claims TE is roughly an entire PS4 cpu (1.6ghz 8 jagcores)
MS claims XSX-a is more powerful then entire X1X cpu. (2.6ghz 8 jagcores)

Microsoft's figure totals a number of audio hardware units, including a programmable unit.

Sony's figure is for the programmable unit only.

Also, MS audio chip is strict to 6 GB of slower ram in XSX

 

[psorcerer]

Except this time is basically watered down version of the same HW, with same feature set, etc.

Doesn't matter.
Each and every rendering effect for example has a cost.
And it's not modular. You cannot disable or enable stuff and expect it to look ok.
You can do resolution changes, but it also not linear.
Essentially your baseline render should look nice on lowest hardware. Which means you cannot make it look much nicer on the highest one.
If there is a need to make games look "next gen" it's pretty impossible to do.

 

[M1chl]

Doesn't matter.
Each and every rendering effect for example has a cost.
And it's not modular. You cannot disable or enable stuff and expect it to look ok.
You can do resolution changes, but it also not linear.
Essentially your baseline render should look nice on lowest hardware. Which means you cannot make it look much nicer on the highest one.
If there is a need to make games look "next gen" it's pretty impossible to do.

I want to state that I agree with developing to 2 different HW targets, like it would with XSX and X1X for example, but for consoles with same HW (feature wise), but different performance targets it's way easier.

Obviously it's going to be tough for MS, since people going to post screens from XSS as XSX screens and vice versa. But I guess it's not that important, while selling more consoles. Epsecially for those prices, which probably next-gen consoles are going to cost.

 

[NEbeast]

It's funny; you see the first page of the thread and is amazing. A positive thread about XsX in GAF and is not being trolled or dismissed? I almost cried, but then I thought: how much it would last?

Then you see the next pages with the introduction of the usual Sony posters and everything changes.

It really makes you think.

All I saw were comparisons being made ,no one was trolling. Xbox fans started baiting/trolling. Anyway, I'm not going to derail the thread.

 

[jimbojim]

Where the heck do those numbers come from? By the most optimistic assessment (comparing max clock to guaranteed clock) it is more than 20% behind.

Wake up, XSeX has 44% more CUs, the gap is MASSIVE. The fact that they managed to cut it in half via OCing is also remarkable on Sony side.

This is your 1.8 TF stronger GPU in numbers. Not bad for a weaker GPU, isn't it.

Triangle rasterisation is 4 triangles per cycle.

PS5:
4 x 2.23 GHz ~ 8.92 Billion triangles per second

XSX:
4 x 1.825 GHz - 7.3 Billion triangles per second

Triangle culling rate is twice number triangles rasterised per cycle.

PS5:
8 x 2.23 GHz - 17.84 Billion triangles per second

XSX:
8 x 1.825 GHz - 14.6 Billion triangles per second

Pixel fillrate is with 4 shader arrays with 4 RBs (render backends) each, and each RB outputtting 4 pixels each. So 64 pixels per cycle.

PS5:
64 x 2.23 GHz - 142.72 Billion pixels per second

XSX:
64 x 1.825 GHz - 116.8 Billion pixels per second

Texture fillrate is based on 4 texture units (TMUs) per CU.

PS5:
4 x 36 x 2.23 GHz - 321.12 Billion texels per second

XSX:
4 x 52 x 1.825 GHz - 379.6 Billion texels per second

Raytracing in RDNA2 is alleged to be from modified TMUs.

PS5:
4 x 36 x 2.23 GHz - 321.12 Billion ray intersections per second

XSX:
4 x 52 x 1.825 GHz - 379.6 Billion Ray intersections per second

 

[Jon Neu]

Doesn't matter.
Each and every rendering effect for example has a cost.
And it's not modular. You cannot disable or enable stuff and expect it to look ok.
You can do resolution changes, but it also not linear.
Essentially your baseline render should look nice on lowest hardware. Which means you cannot make it look much nicer on the highest one.
If there is a need to make games look "next gen" it's pretty impossible to do.

I'm sure MS has a lot of smart people that are capable to design two hardware versions that are pretty scalable between each other while reducing the hassles.

 

[Phrixotrichus]

Except this time is basically watered down version of the same HW, with same feature set, etc.

Still 4tf(that was the latest number right?) seems awfully tight, even for 1080p/30 the moment you figure in next gen effects, especially any form of RT. If visual parity, aside from the resolution/framerate, is a must between those two consoles, and this 4tf machine is supposed to be the "next gen" baseline for the next 5-7 years.......well. In my opinion that smells like a (really stupid)decision coming from a marketing person and not from a developer or engineer......

 

[psorcerer]

I'm sure MS has a lot of smart people that are capable to design two hardware versions that are pretty scalable between each other while reducing the hassles.

Not possible.
The best you can get is the PC situation: game is developed for the "small" hw and then DX12 makes it possible to run on a bigger hw, where you can add resolution/draw distance/tex quality while preserving exactly the same rendering pipeline.

 

[Eliciel]

The key with GP is , how does work its business model? I mean, how much money do get third-parties from the titles they agreed to put in the GP library?

Is it more interesting business-wise for third parties?

I would love to be able to answer that...I doubt that this model "PROMOTES" super high quality games...it handles games as a commodity and some of these commodities are "gamespass sellers"...

I really wonder what they will pay to the developers as well...how will this actually compare to a model were a game cost $60..?

 

[llien]

This is your 1.8 TF stronger GPU in numbers. Not bad for a weaker GPU, isn't it.

(a bunch of meaningless stats)

WTF is this? Aren't there simpler ways to say "but it runs at higher clock"? Jesus.

 

[Jon Neu]

If visual parity, aside from the resolution/framerate, is a must between those two consoles, then this will definitely hold back the higher end one.

So you think they are going to hold back the higher one instead of making the lower one have worse RT?

That's kind of weird.

Not possible.

We will see.

 

[chilichote]

WTF is this? Aren't there simpler ways to say "but it runs at higher clock"? Jesus.

It's not that simple.

 

[RoadHazard]

Now we can officially definitively calculate the raytracing performance of Xbox Series X

XSX - 4 x 52 x 1.825 = 379.6 billion ray triangle intersection per clock

compared to PS5

PS5 - 4 x 36 x 2.23 = 321.12 billion

Surprise, it's the same 18% advantage that the XSX has in raw TF performance over the PS5. Which makes perfect sense, but somehow people didn't think RT also scales with clock speed and that the XSX would have vastly superior RT performance.

 

[longdi]

Microsoft's figure totals a number of audio hardware units, including a programmable unit.

Sony's figure is for the programmable unit only.

Also, MS audio chip is strict to 6 GB of slower ram in XSX

Truth is nobody knows enough, unless Sony gets more transparent.

The only metric comparable is by their jagcore cpu examples.

336gbs is crazy fast just for audio, there is a reason MS partition it this way...
Besides, sorry i doubt PS5 developers will want to eat too much into that 448gbs for audio....

 

[llien]

WTF is this? Aren't there simpler ways to say "but it runs at higher clock"? Jesus.

It's not that simple.

It is... fairly simple, bigger chips tend to run at lower clocks. This includes xx80Ti and Titan.

 

[jimbojim]

I'm not sure how do you know that.
There's no comparison possible as no real numbers exist for NV hw.

I've alwasy thought that 2060 has 24 TF of RT performance, but i was wrong in that.
Anyway, i've fucked it up, sure. I one of my post, maybe at the beginning of this thread, i've said that actually 2060 has bettert RT performance. Overall, Nvidias RT performance surely are better than AMD's

Card               RTX-OPS
RTX 2060           37T
RTX 2060 Super     41T
RTX 2070           45T
RTX 2070 Super     52T
RTX 2080           60T
RTX 2080 Super     63T
RTX 2080 Ti        78T
Titan RTX          84T

 

[jimbojim]

Higher clocks = more heat, worse yield, more cooling needed.

Sony: which worse yields?? We have doubled PS5 production from 5 to 10 million

 

[psorcerer]

I've alwasy thought that 2060 has 24 TF of RT performance, but i was wrong in that.
Anyway, i've fucked it up, sure. I one of my post, maybe at the beginning of this thread, i've said that actually 2060 has bettert RT performance. Overall, Nvidias RT performance surely are better than AMD's

Card               RTX-OPS
RTX 2060           37T
RTX 2060 Super     41T
RTX 2070           45T
RTX 2070 Super     52T
RTX 2080           60T
RTX 2080 Super     63T
RTX 2080 Ti        78T
Titan RTX          84T

"RTX ops" is another nonsense NV number.
It's just some linear combination of various pipeline numbers into one bullshit score.

 

[jimbojim]

RTX ops" is another nonsense NV number.
It's just some linear combination of various pipeline numbers into one bullshit score.

So is the XSX RT numbers are nonsense then.

 

[AphexTwunt]

Xbox Series X 599
Xbox Series S 399
Xbox One S 299

I'm all in for the 599 beast.

 

[psorcerer]

So is the XSX RT numbers are nonsense then.

Nope, it's the only real number we have.
Which is: BVH traversal runs on tex pipeline at 4/cycle
Which gets us to ~380G checks/sec

 

[M1chl]

Lower clocks = less heat, better yield, and less cooling needed.
Higher clocks = more heat, worse yield, more cooling needed.
More CUs at lower frequency = more distributed work so more can be done in parallel, but not done as fast.
Less CUs at higher frequency = each CU can get through more work, but not as much can be done at once.
Less CUs = takes up less space, so die size can be smaller/cheaper or can use the same size but use the extra room for something else. For example IIRC the XB1s die size was bigger than the PS4s despite having a significantly smaller GPU portion because the XB1 had the eSRAM (or eDRAM, can't remember which).

On the deep dive, unsurprisingly the PS5 "secret sauce" Geometry engine also exists here. Shocking! It's almost like they are both made by the same company and that company makes their tech available to everyone and isn't in the business of hiding things.

Would love to see how this isn't a "balanced" design though, since apparently the PS5 is the better designed overall system somehow.

Not sure about that one chief, because yields are mainly concerning the complexity of the chip and if the XSX is bigger chip than, they will suffer more not less. YOu probably going from PC perspective, that clocks are something to do with yields. It's simply, that better chip can do more and with lower temperature, but in PC space you have your line-up and then you scale down the chip and sold it as a lower tier one, so you wont loose all the money.

That's not the case here (well it could be with XSS, tho)...

If Sony is going to have smaller chip, then they can get better yields, period.

 

[Panajev2001a]

Nope, it's the only real number we have.
Which is: BVH traversal runs on tex pipeline at 4/cycle
Which gets us to ~380G checks/sec

It also means that you need to choose between BVH traversal and texture fetch/blend Ops (TMU’s being the Load/Store unit of the GPU kind of).

 

[chilichote]

WTF is this? Aren't there simpler ways to say "but it runs at higher clock"? Jesus.

It is... fairly simple, bigger chips tend to run at lower clocks. This includes xx80Ti and Titan.

Because it's not the same chip only with a higher clock rate. Rather, a chip with fewer CUs, which nevertheless has advantages over a chip with more CUs due to its higher clock rate. Not in all areas but in most of them. And that's why it's exciting to see how this ultimately affects overall performance.

 

[VideodromeX]

It is pretty clear. XSX is the more powerful console. The only metric where the PS5 will trump the XSX is storage IO.

The vagueness and uncertainty are coming from a very select group of warriors looking for secret sauce, similar to what happened with the XBOX ONE at launch. Ignore it... and ignore all warriors.

Someone give this guy a thophy.

 

[Panajev2001a]

Microsoft's figure totals a number of audio hardware units, including a programmable unit.

Sony's figure is for the programmable unit only.

Also, MS audio chip is strict to 6 GB of slower ram in XSX

Exactly, MS’s figure there is like PS3’s claiming 2 TFLOPS because nVIDIA was adding in the equivalent FLOPS from fixed function HW. They quote how many execution units they have for the programmable portion on XSX: 4 FP scalar units for complex operations (logs, exp, trig ops, etc...) and 2x 128 bits wide vector units IIRC which is far less FP throughout than a CU has on RDNA2.

 

[psorcerer]

It also means that you need to choose between BVH traversal and texture fetch/blend Ops (TMU’s being the Load/Store unit of the GPU kind of).

That's pretty obvious, because it's a memory gather in the end.
I don't think NV does anything else here.

 

[MrMiyagi]

Xbox Series X 599
Xbox Series S 399
Xbox One S 299

I'm all in for the 599 beast.

$399 For a 4Tflops Series S sounds like a pretty bad deal if the ps5 DE ends up $399 too.

Even at $299 Series S wouldn't sound like a great option tbh, way too many compromises just to save a measly $100.

 

[M1chl]

Exactly, MS’s figure there is like PS3’s claiming 2 TFLOPS because nVIDIA was adding in the equivalent FLOPS from fixed function HW. They quote how many execution units they have for the programmable portion on XSX: 4 FP scalar units for complex operations (logs, exp, trig ops, etc...) and 2x 128 bits wide vector units IIRC which is far less FP throughout than a CU has on RDNA2.

It's probably first time in history, when we heard about FLOPS of Audio Units. It's great because thanks to Creative Technologies Inc. The audio development in games stagnated greatly.

But I kind of hope that Sony will provide what DSP are used and stuff like that, or maybe it will be digital all the way through, however whole field is really interesting and I hope that we learne more about just the audio unit, not what it can do, we heard that, but how is everything free from noise without errors and that sort of stuff, it's not exactly easy to do audio like this on APU. Because even the DSP side with audio processor in Audio equipment (only digital signal there) is most likely heavy shielded.

And if by any chance I am writing like a retard, I don't doubt them, I just want to learn more. I am kind of audio nerd.

 

[Allandor]

Because it's not the same chip only with a higher clock rate. Rather, a chip with fewer CUs, which nevertheless has advantages over a chip with more CUs due to its higher clock rate. Not in all areas but in most of them. And that's why it's exciting to see how this ultimately affects overall performance.

Clock rate ist not really better if you have the same or better outcome with lower clockrate. Also you have more cache with a more CUs therefore the cache with the lower cu count but higher clocks is faster but also have to make small reads from RAM more often.
Both chips have their own problems with production:
MS:
- lower clocked -> better yield
- bigger chip -> worse yield

Sony:
- higher clocked -> worse yield
- smaller chip -> better yield

Over time both yields should get better.
From what we've seen PS5 seems to have more trouble with production. Therefor the bigger case is needed to keep this thing cool.

btw, sony can't just increase console manufacturing capacity. The companies make contracts for years. Capacities until the end of the year were booked last year. Surprising good yields won't change that much, because the company has still to buy the other components from somewhere (e.g. GDDR6 memory, ...) and thx to covid there is not so much capacity left because some factories just couldn't work.

 

[SoraNoKuni]

There is no doubt Sony will have better audio, it's a really strong field for them and have pioneered so much tech in the field.
They know sound.

I think even hardcore sony fans liked this presentation, but then MS fans just started speaking about the massive victory of the huge buffed titan named XSX and... boom.

I think the only real new data is the ML information, which sounds interesting and may even be something that if Sony lacks could really make a difference in image output.
Other than that if you see my posts I generally trash MS, but I try to be as subjective as I can.

 

[jimbojim]

Clock rate ist not really better if you have the same or better outcome with lower clockrate. Also you have more cache with a more CUs therefore the cache with the lower cu count but higher clocks is faster but also have to make small reads from RAM more often.
Both chips have their own problems with production:
MS:
- lower clocked -> better yield
- bigger chip -> worse yield

Sony:
- higher clocked -> worse yield
- smaller chip -> better yield

Over time both yields should get better.
From what we've seen PS5 seems to have more trouble with production. Therefor the bigger case is needed to keep this thing cool.

btw, sony can't just increase console manufacturing capacity. The companies make contracts for years. Capacities until the end of the year were booked last year. Surprising good yields won't change that much, because the company has still to buy the other components from somewhere (e.g. GDDR6 memory, ...) and thx to covid there is not so much capacity left because some factories just couldn't work.

If capacities were already booked last year, then something changed in the process much later.
Quite the contrary. From 5 to 10 millions increased production isn't bad yields

Sony Boosting Output of PlayStation 5 to Meet Surge in Demand

I think the only real new data is the ML information, which sounds interesting and may even be something that if Sony lacks could really make a difference in image output.

Basically every modern GPUs today have ML. Even PS4 and Pro has that, why PS5 wouldn't.

 

[geordiemp]

Clock rate ist not really better if you have the same or better outcome with lower clockrate. Also you have more cache with a more CUs therefore the cache with the lower cu count but higher clocks is faster but also have to make small reads from RAM more often.
Both chips have their own problems with production:
MS:
- lower clocked -> better yield
- bigger chip -> worse yield

Sony:
- higher clocked -> worse yield
- smaller chip -> better yield

Over time both yields should get better.
From what we've seen PS5 seems to have more trouble with production. Therefor the bigger case is needed to keep this thing cool.

btw, sony can't just increase console manufacturing capacity. The companies make contracts for years. Capacities until the end of the year were booked last year. Surprising good yields won't change that much, because the company has still to buy the other components from somewhere (e.g. GDDR6 memory, ...) and thx to covid there is not so much capacity left because some factories just couldn't work.

LOL, you do realise that MS has the higher clock on the chip at 3.8 GHz ?

The only reason GPU are clocked lower than CPU is they have a longer logic pipeline, and also because at large percentage of the die the cooling technology requirements.

Do you think 2.23 Ghz is high, or 1.825 is low ? What do you think PC cards for RDNA2 will come in at

People talking about yeilds and how final test affects yield parameters is amusing read, you were correct about yields over time though.

 

[llien]

Rather, a chip with fewer CUs, which nevertheless has advantages over a chip with more CUs due to its higher clock rate.

CUs are the "crux" of the GPU.
Assumption, that AMD picked that "else" in a way that will bottlneck the GPU is literally assumption that some engineers out there somewhere are utterly incompetent.

There are zero reasons to believe so, bar console power gap denial.

 

[Aceofspades]

Cerny mentioned that Tempest was running at 64 FLOPS per cycle.

Road to PS5? Or other articles?

 

[M1chl]

Cerny mentioned that Tempest was running at 64 FLOPS per cycle.

It's never like that, because FLOPS are already per second measurements.

 

[llien]

The only reason GPU are clocked lower tha CPU is they have a longer logic pipeline, and also becasue at lareg percentage of the die the cooling technology requirements.

Yeah, and the roots for the pipeline difference are deeper.

There is this wonderful terminology:

Embarrassingly parallel - Wikipedia

en.wikipedia.org

What GPUs are doing is "embarrassingly parallelizable", you can create a looooooong pipeline of work to be done at once and let CUs crunch it. Almost no ifs, all straightforward.

CPUs are exactly the opposite, with wildly random data access and instructions to execute.

 

[Aceofspades]

This is your 1.8 TF stronger GPU in numbers. Not bad for a weaker GPU, isn't it.

Triangle rasterisation is 4 triangles per cycle.

PS5:
4 x 2.23 GHz ~ 8.92 Billion triangles per second

XSX:
4 x 1.825 GHz - 7.3 Billion triangles per second

Triangle culling rate is twice number triangles rasterised per cycle.

PS5:
8 x 2.23 GHz - 17.84 Billion triangles per second

XSX:
8 x 1.825 GHz - 14.6 Billion triangles per second

Pixel fillrate is with 4 shader arrays with 4 RBs (render backends) each, and each RB outputtting 4 pixels each. So 64 pixels per cycle.

PS5:
64 x 2.23 GHz - 142.72 Billion pixels per second

XSX:
64 x 1.825 GHz - 116.8 Billion pixels per second

Texture fillrate is based on 4 texture units (TMUs) per CU.

PS5:
4 x 36 x 2.23 GHz - 321.12 Billion texels per second

XSX:
4 x 52 x 1.825 GHz - 379.6 Billion texels per second

Raytracing in RDNA2 is alleged to be from modified TMUs.

PS5:
4 x 36 x 2.23 GHz - 321.12 Billion ray intersections per second

XSX:
4 x 52 x 1.825 GHz - 379.6 Billion Ray intersections per second

Damn PS5 is monster , add the state of the art I/O and SSD

 

[JLMC469]

Damn PS5 is monster , add the state of the art I/O and SSD

Both are. This SONY fanboy crap has got you blinded. Can’t say anything positive about the competition.

 

[psorcerer]

What GPUs are doing is "embarrassingly parallelizable", you can create a looooooong pipeline of work to be done at once and let CUs crunch it. Almost no ifs, all straightforward.

That's not correct anymore.
All the newer "next gen" effects are much less parallel, just by virtue of being less local.
Reflections, GI, etc.

 

[YoodlePro]

This is your 1.8 TF stronger GPU in numbers. Not bad for a weaker GPU, isn't it.

Triangle rasterisation is 4 triangles per cycle.

PS5:
4 x 2.23 GHz ~ 8.92 Billion triangles per second

XSX:
4 x 1.825 GHz - 7.3 Billion triangles per second

Triangle culling rate is twice number triangles rasterised per cycle.

PS5:
8 x 2.23 GHz - 17.84 Billion triangles per second

XSX:
8 x 1.825 GHz - 14.6 Billion triangles per second

Pixel fillrate is with 4 shader arrays with 4 RBs (render backends) each, and each RB outputtting 4 pixels each. So 64 pixels per cycle.

PS5:
64 x 2.23 GHz - 142.72 Billion pixels per second

XSX:
64 x 1.825 GHz - 116.8 Billion pixels per second

Texture fillrate is based on 4 texture units (TMUs) per CU.

PS5:
4 x 36 x 2.23 GHz - 321.12 Billion texels per second

XSX:
4 x 52 x 1.825 GHz - 379.6 Billion texels per second

Raytracing in RDNA2 is alleged to be from modified TMUs.

PS5:
4 x 36 x 2.23 GHz - 321.12 Billion ray intersections per second

XSX:
4 x 52 x 1.825 GHz - 379.6 Billion Ray intersections per second

I think it doesn't matter that you bold it, make it red, put max font or paint them on a pair of huge titties or some Cardi B ass, people will still ignore these numbers and push the fake narrative of OMG the difference between these GPUS is massssssiiiive

 

[TBiddy]

Road to PS5? Or other articles?

PlayStation 5 uncovered: the Mark Cerny tech deep dive

On March 18th, Sony finally broke cover with in-depth information on the technical make-up of PlayStation 5. Expanding …

www.eurogamer.net

It's never like that, because FLOPS are already per second measurements.

See above.

Damn PS5 is monster , add the state of the art I/O and SSD

You're right it is, but take it somewhere else.

 

[DonMigs85]

kinda reminds me of the Pentium 3/Celeron derivative in OG Xbox only having 128K L2 cache

 

[YoodlePro]

Clock rate ist not really better if you have the same or better outcome with lower clockrate. Also you have more cache with a more CUs therefore the cache with the lower cu count but higher clocks is faster but also have to make small reads from RAM more often.
Both chips have their own problems with production:
MS:
- lower clocked -> better yield
- bigger chip -> worse yield

Sony:
- higher clocked -> worse yield
- smaller chip -> better yield

Over time both yields should get better.
From what we've seen PS5 seems to have more trouble with production. Therefor the bigger case is needed to keep this thing cool.

btw, sony can't just increase console manufacturing capacity. The companies make contracts for years. Capacities until the end of the year were booked last year. Surprising good yields won't change that much, because the company has still to buy the other components from somewhere (e.g. GDDR6 memory, ...) and thx to covid there is not so much capacity left because some factories just couldn't work.

Well you're missing the fact that Sony has its own factories. For example the one in Japan churning out a PS4 every 30 seconds. If they start repurposing that for PS5?

 

[M1chl]

PlayStation 5 uncovered: the Mark Cerny tech deep dive

On March 18th, Sony finally broke cover with in-depth information on the technical make-up of PlayStation 5. Expanding …

www.eurogamer.net

See above.

You're right it is, but take it somewhere else.

HMMM very strange.