Support NeoGAF

[TLZ]

Zen 3000 with 8 cœurs / 16 threads @ 3,2 GHz
Radeon Navi 72 cu (36x2), 4608 shaders units @ 1550 MHz
14,2 TFLOPS
16 Go / 24 GB of GDDR6
Hardware ray tracing
2TB SSD

Latest Sony PlayStation 5 Rumored Specs Boast Powerful AMD Radeon Navi And Ryzen Tech, 2TB SSD

For starters, it's claimed that there will be an 8-core/16-thread Ryzen 3000 processor clocked at 3.2GHz.

hothardware.com

Follow the source. "Insider" never named. May as well be a pastebin imo. Specs are on the very optimistic end even for 499, even though possible.

PlayStation 5: primi dettagli hardware pubblicati da un insider

Un insider ha appena messo in rete quelle che potrebbero essere le caratteristiche tecniche e hardware di PlayStation 5 di Sony

translate.google.com

PlayStation 5 / PS5 - Pastebin.com

Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.

pastebin.com

 

[StreetsofBeige]

beast confirmed. it has to be euv right SonGoku

Pastebin sent me a message that all next gen pasties are done by leaker Koguson!

 

[TeamGhobad]

Pastebin sent me a message that all next gen pasties are done by leaker Koguson!

400mm^2 on euv is possible but even 60-64 CUs is beast.

 

[ArabianPrynce]

Pastebin sent me a message that all next gen pasties are done by leaker Koguson!

wait wat

 

[DeeDogg_]

Zen 3000 with 8 cœurs / 16 threads @ 3,2 GHz
Radeon Navi 72 cu (36x2), 4608 shaders units @ 1550 MHz
14,2 TFLOPS
16 Go / 24 GB of GDDR6
Hardware ray tracing
2TB SSD

Latest Sony PlayStation 5 Rumored Specs Boast Powerful AMD Radeon Navi And Ryzen Tech, 2TB SSD

For starters, it's claimed that there will be an 8-core/16-thread Ryzen 3000 processor clocked at 3.2GHz.

hothardware.com

FIVE HUNDRED AND NINETY NINE YOU ESS DOLLARS

 

[molly14]

so basically a 2080ti in a console astonishing ,please be true,and 499 pounds with Sony taking a 200 pound loss on each sale ,only way this is possible considering just the 2080 ti retails over 700 pounds by itself here

 

[StreetsofBeige]

so basically a 2080ti in a console astonishing ,please be true,and 499 pounds with Sony taking a 200 pound loss on each sale ,only way this is possible considering just the 2080 ti retails over 700 pounds by itself here

Anyone know what the ballpark manufacturer cost is for GPUs?

- So let's say a new GPU selling at stores is $500 retail
- Retail store's cost (manufacturer selling price to a store) is $XXX
- Console/PC maker's cost (manufacturer selling price to another company) is $XXX
- Manufacturer's cost to make GPU is $XXX

I understand there will be wholesaler costs in here too supplying retailers, but let's skip that for simplicity.

 

[ArabianPrynce]

Anyone know what the ballpark manufacturer cost is for GPUs?

- So let's say a new GPU selling at stores is $500 retail
- Retail store's cost (manufacturer selling price to a store) is $XXX
- Console/PC maker's cost (manufacturer selling price to another company) is $XXX
- Manufacturer's cost to make GPU is $XXX

I understand there will be wholesaler costs in here too supplying retailers, but let's skip that for simplicity.

who is that leaker you mentioned?

 

[Thedtrain]

who is that leaker you mentioned?

The great Koguson

 

[ArabianPrynce]

The great Koguson

Which forum is he from and how realiable is he. 14tf on the ps5 is a bit ridiculous considering what we know of navi in regards to die size and power consumption. He could be talking about 14tf in terms of gcn not navi.

 

[Burrito Bandito]

Just make it 599,- and deliver a beast, Sony. If people cant afford it they could easily wait or buy the PS4 Slim or Pro.
Its gonna fly off the shelves anyway. Be it 399 (Forget it!) 499 (My prediction.) or 599 (Would still buy it). Gamers on PC pay for mid range cards way over 300$/€ and thats just the GPU alone. I think Sony can ask for a 599,- console without getting as much flak as they did back in 2006.

with as much market share that they have i would be utterly shocked if they took any sort of risk or loss per console. They will make a &450 console and sell it for $499.

MS is the only one that makes any sense to take a loss on the console. i could see them making a $550 or $600 console and selling it for $499. I just dont see the sense in Sony doing that.

 

[bhunachicken]

I still reckon the machine will be around the 1080 TI ballpark (~11TF).

There will be those who say that's too old and too slow, but devs will be able to heavily optimise and get better performance out of it than a PC does.

60FPS @ 4K will be easily possible on it. Plus, there will be the Secret Sauce (probably hybrid raytracing) to make things juicer.

Prices?
$499 US
£445 GBP
499 EUR

 

[FranXico]

Which forum is he from and how realiable is he. 14tf on the ps5 is a bit ridiculous considering what we know of navi in regards to die size and power consumption. He could be talking about 14tf in terms of gcn not navi.

 

[DemonCleaner]

the leak reads like someone who read gaf/ree next gen threads

gaf -> internets -> gaf

scientists call it "the gaf circle"

 

[TeamGhobad]

anything below 60CUs is a disappointment. but i think they will land on 56.

 

[pawel86ck]

According to SONY PS5 is made for HARDCORE gamers, but what does it really mean? PS5 will be expensive and really fast?

 

[TeamGhobad]

According to SONY PS5 is made for HARDCORE gamers, but what does it really mean? PS5 will be expensive and really fast?

it means another 599 system. MS could swoop in with a 400 system and win.

 

[THE:MILKMAN]

According to SONY PS5 is made for HARDCORE gamers, but what does it really mean? PS5 will be expensive and really fast?

Isn't the 'for hardcore gamers' line from an opinion piece by the WSJ?

 

[pawel86ck]

it means another 599 system. MS could swoop in with a 400 system and win.

At 599$ I expect 12TF+ Navi GPU, not 8TF. I bought PS3 at launch for 599 euro and I'm willing to pay 599 euro yet another time now for PS5 if hardware will be impressive.

Isn't the 'for hardcore gamers' line from an opinion piece by the WSJ?

Here you can read about it

https://www.gamesradar.com/amp/ps5-...-the-latest-features-according-to-new-report/

 

[TeamGhobad]

At 599$ I expect 12TF+ Navi GPU, not 8TF. I bought PS3 at launch for 599 euro and I'm willing to pay 599 euro yet another time now for PS5 if hardware will be impressive.

i dont think 600bucks for a console you will use for 8-9years is a bad deal. but unfortunately the market disagrees with us. I personally believe the PS3 failed because it had no games and multiplatform games looked better on xbox 360.

 

[ANIMAL1975]

Beast? more like KAIJU category V
Assuming leak its legit yeah EUV is the only way to fit 72CU (enabled) while keeping die size reasonable ~390-400mm2

For those worried $500 launch price might affect sales *cough* ArabianPrynce *cough*
The reason PS3 did so bad is that it took 3 years to match the 360 launch price, its impressive how much brand recognition carried PS3 early years

Spoiler

Embrace yourselves

FIVE HUNDRED AND NINETY NINE YOU ESS DOLLARS

That's what more than one leak says, 599 with Sony taking an 100 loss for Kaiju.

who is that leaker you mentioned?

Pastebin Cell lite for RT leak.

with as much market share that they have i would be utterly shocked if they took any sort of risk or loss per console. They will make a &450 console and sell it for $499.

MS is the only one that makes any sense to take a loss on the console. i could see them making a $550 or $600 console and selling it for $499. I just dont see the sense in Sony doing that.

Even Penello thinks otherwise about MS be willing to take a loss (is words on the previous next gen OT in Era). As for Sony, "with as much market share that they have" and paying plus subscribers, i think they're the ones that make more sense for taking initial losses per machine.

According to SONY PS5 is made for HARDCORE gamers, but what does it really mean? PS5 will be expensive and really fast?

More expensive than PS4 yes, but not PS3 levels of expensive,... and really fast yes.

 

[THE:MILKMAN]

Here you can read about it

https://www.gamesradar.com/amp/ps5-...-the-latest-features-according-to-new-report/

Yeah I have checked them out and IMO I don't really trust the opinion or claims of inside sources of the WSJ article writer.

 

[bitbydeath]

According to SONY PS5 is made for HARDCORE gamers, but what does it really mean? PS5 will be expensive and really fast?

Cerny said the price would be well received despite the hardware being super powerful.

They’ll no doubt be selling at a loss, probably with the subscriptions and a game in mind to offset the costs.

 

[pawel86ck]

i dont think 600bucks for a console you will use for 8-9years is a bad deal. but unfortunately the market disagrees with us. I personally believe the PS3 failed because it had no games and multiplatform games looked better on xbox 360.

I dont remember people in my country who complained about PS3 price. In fact I personally considered 599 euro cheap because it was very hard to buy a decent PC for 599 euro (I have paid 4x as much for my PC back then). But I do remember people complainig about lack of good quality games on PS3. Motorstorm was good, resistance was average, but that wasnt good enough to convince some people to pay more for ps3 and especially when the majority of multiplatform games run worse compared to x360.

IMO lack of good games were a real problem and not a price. If PS3 would get god of war 3, GT5, Metal gear solid 4 at launch PS3 would sell very well even at 599$. But since xbox 360 had more games in 2007 Sony was forced to reduce the price.

IMO If Sony will deliver 12TF+ Navi beast and launch their console with good games, PS5 will sell very good even at 599$ price point. Just look at smartphone prices, people are willing to pay 700-1000$ every year for a little bit better phone, so paying 599$ for a 6-8 year gaming perspective is still cheap price, and especially if you consider how expensive it is to build high end PC. 2080ti alone cost 1200$, so twice as much as 599$.

 

[THE:MILKMAN]

Cerny said the price would be well received despite the hardware being super powerful.

They’ll no doubt be selling at a loss, probably with the subscriptions and a game in mind to offset the costs.

Cerny was quoted by Peter Rubin as saying:

I believe that we will be able to release it at an SRP [suggested retail price] that will be appealing to gamers in light of its advanced feature set.

This to my mind is a PR 101 pre-prepared statement given to Mark by his PR team to answer the inevitable price question from Peter Rubin.

It means nothing/everything/anything you want.

 

[sinnergy]

I dont remember people in my country who complained about PS3 price. In fact I personally considered 599 euro cheap because it was very hard to buy a decent PC for 599 euro (I have paid 4x as much for my PC back then). But I do remember people complainig about lack of good quality games on PS3. Motorstorm was good, resistance was average, but that wasnt good enough to convince some people to pay more for ps3 and especially when the majority of multiplatform games run worse compared to x360.

IMO lack of good games were a real problem and not a price. If PS3 would get god of war 3, GT5, Metal gear solid 4 at launch PS3 would sell very well even at 599$. But since xbox 360 had more games in 2007 Sony was forced to reduce the price.

IMO If Sony will deliver 12TF+ Navi beast and launch their console with good games, PS5 will sell very good even at 599$ price point. Just look at smartphone prices, people are willing to pay 700-1000$ every year for a little bit better phone, so paying 599$ for a 6-8 year gaming perspective is still cheap price, and especially if you consider how expensive it is to build high end PC. 2080ti alone cost 1200$, so twice as much as 599$.

I bought it day one and the 599 euro felt like a complete ripoff imo. And the launch games looked worse or on par with Xbox 360.

It also sold bad, that the needed to price cut and introduce new models ..

599 is just bad, 399 is the sweet spot, and 499 is just a spot that the first hardcore will want to spend.

 

[bitbydeath]

Cerny was quoted by Peter Rubin as saying:


This to my mind is a PR 101 pre-prepared statement given to Mark by his PR team to answer the inevitable price question from Peter Rubin.

It means nothing/everything/anything you want.

Why do you think that?
It looks very specific to me.

 

[Burrito Bandito]

yall should check your expectations if you think Sony is selling at a $100 loss lol. theres zero business sense in that for them

 

[bitbydeath]

yall should check your expectations if you think Sony is selling at a $100 loss lol. theres zero business sense in that for them

That’s very common for the beginning of a generation.

PS3 was even more bonkers with a $300 loss on each console sold. (Don’t expect that again.)

 

[THE:MILKMAN]

Why do you think that?
It looks very specific to me.

It is how it reads to me. Also Mark is chief architect and not someone that would usually comment on price but reading the Wired article it is clear other Sony reps, no doubt including PR, were in the room and likely the source of that line about price. The biggest surprise with all this is that Peter left this out of the original article and seemed to scramble to post it on Twitter when the forums and Twitter screamed $599!! and Sony presumably panicked and got Peter to quickly post that quote! (only half joking).

All IMO of course.

 

[SpinningBirdKick]

anything below 60CUs is a disappointment. but i think they will land on 56.

I think MS will probably land on 60 active CUs.

It will be disappointing.

 

[DemonCleaner]

thought of the day on PS5 / navi / RDNA:

some weeks back when DF's richard did his next gen speculation video, we uniformely had the opion it's just a summary what we had speculated in this thread over the last half year. yet he noticed something that everyone else missed. transistor density on Navi10 did increase significantly over the last GCN installments. if we actually do the math it accounts to 63% over the last somewhat comparable polaris 10 die with 36CUs as i tried to show in the following chart.

*note that is total transistors on die divded by the CU count of a full (undisabled) die

so what are all those additinal transistors for? as we learned in the AMD presentation at E3 they "streamlined" the graphics pipeline. historically nvidia had had a longer (and therfore more die space intensive) pipeline which allowed them to clock relatively high compared to amd on the same node. now that alone can't have been the driving force behind amd's redesign as we barely - if at all - see better clocks compared to vega on 7nm. so what do you need all the extra die space for? it surely does not simply go to waste.

well ok, you could say at this point that vega already had quite a bump in complexety without seeing any significant real world gains. on that topic keep in mind, firstly: vega 10 and 20 have different proportions between frontend components and CU count plainly because it has 64CUs and not around 40. secondly: vega already had a rework of the pipeline which should have enabled a feature called "primitive shaders" that never came to fruition but needed a serious rework of the CUs.

so what then is the kicker behind all this added complexity?

well it might have somehting to do with the fixed function/shader hybrid approach to raytracing, which became public through the AMD patent a few days ago:

https://www.techpowerup.com/256975/amd-patent-shines-raytraced-light-on-post-navi-plans

the authors seem to stress, that this approach is in many ways superior than going with all fixed funcition hardware. especially that it minimizes the die area that lies idle during the other rendering steps compared to fully dedicated hardware.

the ability to use the whole of your available shaders for ray shooting and not rely on just a small part of your die for that step hopefully also alleviates or bypasses the necessity of a denoising step which there is seamingly no dedicated hardware on the Navi die.

to implement that method on die, you would not only need the BVH hardware ("ray intersection engine" as they call it in the patent) but also need to rework the CUs and add paths between components.

so that might mean, that a big chunk of that extra complexity we are seeing might be the bet on what's the next big thing in realtime graphics for the coming decade: hybrid rendering.

 

[JohnnyFootball]

That’s very common for the beginning of a generation.

PS3 was even more bonkers with a $300 loss on each console sold. (Don’t expect that again.)

It was common prior to last gen, but I recall that with this gen Sony and MS said no more.

 

[ArabianPrynce]

The great Koguson

Why am I getting trolled lol

 

[Panajev2001a]

thought of the day on PS5 / navi / RDNA:

some weeks back when DF's richard did his next gen speculation video, we uniformely had the opion it's just a summary what we had speculated in this thread over the last half year. yet he noticed something that everyone else missed. transistor density on Navi10 did increase significantly over the last GCN installments. if we actually do the math it accounts to 63% over the last somewhat comparable polaris 10 die with 36CUs as i tried to show in the following chart.

*note that is total transistors on die divded by the CU count of a full (undisabled) die

so what are all those additinal transistors for? as we learned in the AMD presentation at E3 they "streamlined" the graphics pipeline. historically nvidia had had a longer (and therfore more die space intensive) pipeline which allowed them to clock relatively high compared to amd on the same node. now that alone can't have been the driving force behind amd's redesign as we barely - if at all - see better clocks compared to vega on 7nm. so what do you need all the extra die space for? it surely does not simply go to waste.

well ok, you could say at this point that vega already had quite a bump in complexety without seeing any significant real world gains. on that topic keep in mind, firstly: vega 10 and 20 have different proportions between frontend components and CU count plainly because it has 64CUs and not around 40. secondly: vega already had a rework of the pipeline which should have enabled a feature called "primitive shaders" that never came to fruition but needed a serious rework of the CUs.

so what then is the kicker behind all this added complexity?

well it might have somehting to do with the fixed function/shader hybrid approach to raytracing, which became public through the AMD patent a few days ago:

https://www.techpowerup.com/256975/amd-patent-shines-raytraced-light-on-post-navi-plans

the authors seem to stress, that this approach is in many ways superior than going with all fixed funcition hardware. especially that it minimizes the die area that lies idle during the other rendering steps compared to fully dedicated hardware.

the ability to use the whole of your available shaders for ray shooting and not rely on just a small part of your die for that step hopefully also alleviates or bypasses the necessity of a denoising step which there is seamingly no dedicated hardware on the Navi die.

to implement that method on die, you would not only need the BVH hardware ("ray intersection engine" as they call it in the patent) but also need to rework the CUs and add paths between components.

so that might mean, that a big chunk of that extra complexity we are seeing might be the bet on what's the next big thing in realtime graphics for the coming decade: hybrid rendering.

I do not think Navi 10 / this July’s Navi has Hybrid RT, one thing Richard L. missed in his analysis was that the shader units doubled in width (vector units went 32 wide from 16 of GCN and L0/L1 caches and many internal busses grew quite a bit too).

 

[LordOfChaos]

Zen 2 yields over 85%: TSMC is doing a very good job

According to our sources, the yields of Zen 2 are enhancing quickly. Three months ago we wrote that Zen 2 yields (100% fully working chip) were about 70%. Today, yields are about 85%!Zen 2 yields are still a little bit lower than Zeppelin ones (Today at +90%), but at this rate Zen 2 will be...

www.bitsandchips.it

TSMC N7 is settling down fast

 

[ethomaz]

DAT confidence in own product

 

[ethomaz]

Zen 2 yields over 85%: TSMC is doing a very good job

According to our sources, the yields of Zen 2 are enhancing quickly. Three months ago we wrote that Zen 2 yields (100% fully working chip) were about 70%. Today, yields are about 85%!Zen 2 yields are still a little bit lower than Zeppelin ones (Today at +90%), but at this rate Zen 2 will be...

www.bitsandchips.it

TSMC N7 is settling down fast

Not bad for a new process... the big issue is 7nm wafers being twice expensive than 16nm wafers.

 

[ArabianPrynce]

Not bad for a new process... the big issue is 7nm wafers being twice expensive than 16km wafers.

can you explain what this means in laymen terms?

 

[LordOfChaos]

Not bad for a new process... the big issue is 7nm wafers being twice expensive than 16km wafers.

Well yeah, it was just one big transistor per city!

 

[TeamGhobad]

loosely nexgen related. have we seen any 4k benchmarks using navi?

 

[DeepEnigma]

DAT confidence in own product

Or they baited them into pricing. Come out high, because you know nVidia will no doubt match it. They love high prices.

Then undercut. It’s funny, because what they cut the price to right now, is around what gamers all expected before the E3 pricing that came in a hair higher.

 

[JohnnyFootball]

DAT confidence in own product

Or they baited them into pricing. Come out high, because you know nVidia will no doubt match it. They love high prices.

Then undercut. It’s funny, because what they cut the price to right now, is around what gamers all expected before the E3 pricing that came in a hair higher.

It's rare that something can be spun so beautifully one way or the other. Well done AMD. Well done Nvidia.

A small win for gamers.

 

[DemonCleaner]

I do not think Navi 10 / this July’s Navi has Hybrid RT,

im not saying that NAVI 10 has hybrid RT. they've shown as much in their roadmap. im speculating that if it comes with RDNA2/Consoles they already might have layed the groundwork for that in their CU and frontend redesign and won't start over with RDNA2. hence the higher complexity. would at least make somwhat sense, wouldn't it?

one thing Richard L. missed in his analysis was that the shader units doubled in width (vector units went 32 wide from 16 of GCN and L0/L1 caches and many internal busses grew quite a bit too).

cashes grew, but im not sure the rest is accurate. you have any source for that? CUs are still 64 lanes wide. shouldn't total buswidth per CU therefore be unchanged?


see GCN:

Navi:

 

[Panajev2001a]

im not saying that NAVI 10 has hybrid RT. they've shown as much in their roadmap. im speculating that if it comes with RDNA2/Consoles they already might have layed the groundwork for that in their CU and frontend redesign and won't start over with RDNA2. hence the higher complexity. would at least make somwhat sense, wouldn't it?

Sure, but it would be an expensive gamble on the PC space unless you are super confident about your die failure rate as you are making a much bigger chip than you need that year, but yeah not impossible .

cashes grew, but im not sure the rest is accurate. you have any source for that? CUs are still 64 lanes wide. shouldn't total buswidth per CU therefore be unchanged?


see GCN:

Navi:

AMD Announces Radeon RX 5700 XT & RX 5700: The Next Gen of AMD Video Cards Starts on July 7th At $449/$379

www.anandtech.com

Doubling the width of each vector unit from 16 wide SIMD to 32 wide SIMD, 64 was just the size of the threads batch size where they would take 4 cycles on GCN and now 2 cycles on Navi... that is the legacy 64 wide wavefront size, the new wavefront size is 32 that guarantees single cycle processing on Navi and a lot better dynamic branching efficientcy as you throw away a lot less work per wave and you can have diverging waves.

 

[LordOfChaos]

A Look At The AMD Zen 2 Core

Ahead of the highly anticipated Ryzen 3000 desktop series launch, here is a look at the AMD Zen 2 core microarchitecture.

fuse.wikichip.org

 

[bhunachicken]

It would be a tremendous fuck up if Sony were to price the machine above £449.

£399 would be great, but I imagine the losses on that would be a smidge too high. So, £449 it is.

I'm already putting pound coins into a jar on daily basis, so I'll have 449 of 'em by launch

 

[Aceofspades]

It would be a tremendous fuck up if Sony were to price the machine above £449.

£399 would be great, but I imagine the losses on that would be a smidge too high. So, £449 it is.

I'm already putting pound coins into a jar on daily basis, so I'll have 449 of 'em by launch

They would outsell MS even if they priced it $800

Edit: I mean MS should be more price sensitive than Sony since Xbox doesn't carry the same brand power of PlayStation.

 

[Insane Metal]

Well, we now know that Zen 2 is better at gaming then Zen 1. Anyway, here's a nice preview:

AMD Ryzen 9 3900X 12 Core and Ryzen 7 3700X 8 Core CPU Review Leaks Out – Gaming Performance On Par With Intel’s 9th Gen, Ryzen 9 3900X Consumes Less Power Than i9-9900K

 

[DemonCleaner]

Sure, but it would be an expensive gamble on the PC space unless you are super confident about your die failure rate as you are making a much bigger chip than you need that year, but yeah not impossible .

well the whole point of the patent is, that it wouldn't waste tooooo much die space [yeah i realize that im somewhat contradicting myself here ]. maybe also the reason why we not seeing big Navi just yet.

Doubling the width of each vector unit from 16 wide SIMD to 32 wide SIMD, 64 was just the size of the threads batch size where they would take 4 cycles on GCN and now 2 cycles on Navi... that is the legacy 64 wide wavefront size, the new wavefront size is 32 that guarantees single cycle processing on Navi and a lot better dynamic branching efficientcy as you throw away w lot less work per wave and you can have diverging waves.

now that you put it that way, some of the slides make much more sense. but after taking a look at the die render im not sure if the concerning components could really contribute the bulk to the need for 63% or even 32% (vs. vega 10) more transistors.

maybe i will take a closer look at that tommorow.