Support NeoGAF

[onQ123]

Ah okay, gotcha' . So it could be for any situation, including potentially a future PS5, etc.

Yeah

 

[onQ123]

PS4 Pro can't have Puma CPU because Puma doesn't support HSA

 

[Negotiator]

PS4 Pro can't have Puma CPU because Puma doesn't support HSA

Who knows, maybe it's a modified Puma, like the semi-custom Radeon 7850.

 

[TheThreadsThatBindUs]

There's literally no reason for the PS4 Pro CPU (and by extension the XB1X) CPU to be anything other than a stock Jaguar, optimized for a more modern manufacturing process and clocked higher.

Let's not forget that the entire point of these mid-gen updates is for them to be 100% compatible with their predecessor consoles within the same family.

I consider it extremely unlikely that there are any architectural improvements to the Jaguar cores in the Pro (or XB1X), as any such changes would hold the potential to break compatibility—i.e. the core design goal—for the sake of whatever incremental improvement that would never even be utilized (i.e. these boxes will be stuck playing PS4/XB1 games for the entirety of their existence).

 

[onQ123]

Who knows, maybe it's a modified Puma, like the semi-custom Radeon 7850.

They talk about turning off extra hardware features on the new CPU

5. The method of claim 1, wherein executing the application on the new CPU includes executing the application on the new CPU with selected features of the new CPU that are not present on legacy CPU disabled.

6. The method of claim 5, wherein executing the application on the new CPU with selected features of the new CPU that are not present on the legacy CPU disabled includes disabling a dedicated loop predictor of the new CPU when the new CPU and its associated prediction logic include the dedicated loop predictor, but the legacy CPU does not.

7. The method of claim 5, wherein executing the application on the new CPU with selected features of the new CPU that are not present on legacy CPU disabled includes disabling a micro-op cache of the new CPU when the new CPU includes the micro-op cache, but the legacy CPU does not.

 

[Panajev2001a]

They talk about turning off extra hardware features on the new CPU

5. The method of claim 1, wherein executing the application on the new CPU includes executing the application on the new CPU with selected features of the new CPU that are not present on legacy CPU disabled.

6. The method of claim 5, wherein executing the application on the new CPU with selected features of the new CPU that are not present on the legacy CPU disabled includes disabling a dedicated loop predictor of the new CPU when the new CPU and its associated prediction logic include the dedicated loop predictor, but the legacy CPU does not.

7. The method of claim 5, wherein executing the application on the new CPU with selected features of the new CPU that are not present on legacy CPU disabled includes disabling a micro-op cache of the new CPU when the new CPU includes the micro-op cache, but the legacy CPU does not.

Pretty sure Jaguar or any Jaguar related evolution does not have a loop predictor or especially the u-ops cache (Ryzen features).

 

[onQ123]

Pretty sure Jaguar or any Jaguar related evolution does not have a loop predictor or especially the u-ops cache (Ryzen features).

Yeah I'm sure they are talking about a different CPU



[0029] In certain aspects of the present disclosure, in BC mode the size of the retirement queues 132, the size of the scheduling queues 136, or the size of the SIMD 142 or GP register banks 144 of the CPU may be reduced to match, or to more closely approximate, their respective size for the legacy CPU. To be clear, this reduction takes the form of reducing the usable portion of the resource, e.g. restricting the number of physical registers available to the application in BC mode; the full register bank would be available for use by applications when not in BC mode.

[0030] The execution units 150 typically include SIMD pipes 152 that perform a number of parallel operations on multiple data fields contained in 128-bit or wider SIMD registers contained in the SIMD register bank 142, arithmetic and logic units (ALUs) 154 that perform a number of logical, arithmetic, and miscellaneous operations on GPRs contained in the GP register bank 144, and address generation units (AGUs) 156 that calculate the address from which memory should be stored or loaded. There may be multiple instances of each type of execution unit, and the instances may have differing capabilities, for example a specific SIMD pipe 152 may be able to perform floating point multiply operations but not floating point add operations.

[0031] In certain aspects of the present disclosure, in BC mode the usable number of ALUs, AGUs or SIMD pipes may be reduced to match, or to more closely approximate, the respective number of such units that exist on the legacy CPU.

 

[Lady Gaia]

It’s certainly intriguing to see them actively exploring the space. Whether they actually follow through with compatibility or just lay a patent minefield for a competitor is less clear.

 

[onQ123]

It’s certainly intriguing to see them actively exploring the space. Whether they actually follow through with compatibility or just lay a patent minefield for a competitor is less clear.

It might have been some truth to the PS4 Pro with a different CPU, it might have been in the testing stage but got cut.

 

[TheThreadsThatBindUs]

It’s certainly intriguing to see them actively exploring the space. Whether they actually follow through with compatibility or just lay a patent minefield for a competitor is less clear.

Would this patent and others like it make a difference? Especially when MS' solution will primarily be driven by the fact that their software platform is virtualized on XB1?

 

[AmyS]

TSMC Updates its Silicon Menu

First 7-nm chips, EUV migration described

SANTA CLARA, Calif. — TSMC reported progress in 7 nm and extreme ultraviolet (EUV) lithography and bolstered a planar process that competes with fully depleted silicon-on-insulator at an annual event here. It also gave updates on its work in packaging and platforms for key market segments.

TSMC sketched out what it called a relatively simple process of porting design rules and IP to an N7+ process using EUV that it could put into production in 2019. The process can deliver 20% greater density, 8–10% higher speeds, or 15–20% less power than its current N7 node. Compared to its 16FFC process, N7+ can enable 30% higher speed or 50% less power on an ARM A72 core, said Cliff Hou, vice president of R&D for design technology at TSMC.

The foundry will provide a utility to port immersion design rules to the EUV process that will ”clean up most of the layout differences," said Hou. Overall, the work of moving from N7 to N7+ should represent about a third of the effort of migrating to a new node, he added.

And on stacked DRAM (HBM2).

In packaging, TSMC said that it is working on a new variant of InFO, its wafer-level fab-out technique famously used in Apple's latest A Series processors. InFO-MS will integrate logic and memory and is first being targeted for use with the latest high bandwidth memory (HBM2) in efforts among TSMC, Samsung, and SK Hynix.

Separately, Open-Silicon announced Tuesday that it has validated for use in system-in-package designs its HBM2 IP subsystem made using TSMC's 16-nm process and its CoWoS 2.5-D chip stacking technology. It supports data rates up to 2 Gbits/second per pin. The company expects that a 7-nm version will hit 2.4 Gbits/s.

”Probably no company in the world has done as much to take us into 3D chip stacks as TSMC," said Wally Rhines, chief executive for Mentor Graphics, which provides verification software for some of the foundry's stacks.

http://www.eetimes.com/document.asp?doc_id=1332293

Hopefully, the chips in PS5 (and the future Xbox) can take advantage of all these things, and no reason they couldn't, since already PS4 Pro and XBX1 SoC are both fabbed at TSMC.

 

[TheThreadsThatBindUs]

Global Foundries are looking even better:

https://www.globalfoundries.com/sites/default/files/product-briefs/7lp-product-brief.pdf

 

[Dehnus]

http://gamingbolt.com/ps5-will-like...-will-possibly-launch-in-2019-michael-pachter

I don't know if Pachter is getting wiser or I'm getting dumber but I agree with him.

I can see the PS5 do a minor jump to reach 4K as a standard but most effort be pushed towards CPU to improve performance for PSVR(2).

Half close if old.

Well, actually it isn't that strange. Imagine the difference between Xbox one and One x. GPU wise it's like a factor 4.5. But that is an extreme, in a way the Xbox One X with a beefier CPU would be a new gen if they let go of the compatibility with old systems and allow Xbox one X only games.

Now imagine that difference with a PS 4 Pro and ordinary PS 4. By just putting in a beefier CPU and letting go of the older systems (so games are no longer forced to work on PS4 systems) you suddenly have quite a beefy system at hand. 4.2 teraflops / 1.84 = 2.3. This would mean the PS5 would be a system edging on 10 teraflops. Add to that a beefier CPU that can push a lot more data to the GPU, letting go of forced PS4 versions? And you suddenly have a new generation.

I mean it's a system that is more than 5 times as powerful as an ordinary PS4 on GPU alone. In comparison the Xbox 360 vs the Xbox one is a factor of 5 in flop count difference, the PS3 to PS 4 about the same.

A 10 tf system with a Ryzen based CPU and a good 24GB of GDDR5X of memory would go pretty far in 2019. (AKA a Ravenridge specialized APU). Backwards compatibility as well, and MS would have an issue, in competing big time.

 

[TheThreadsThatBindUs]

MS will have an issue competing with PS5, regardless of hardware, unless they do something about their first party exclusives portfolio.

 

[Bitch Pudding]

MS will have an issue competing with PS5, regardless of hardware, unless they do something about their first party exclusives portfolio.

Will be interesting, most likely scenario for now is Sony getting at least a one year head start, which on the other hand gives MS the opportunity to come up with a more powerful one which they could (and most likely will) announce before PS5 launches.

 

[ayob]

I see ps5 being a full step with ps4 backwards compatibility. It'll be "the first" next gen 4k box, probably twice or four times the power of the Xbox one x

Edit: release March 2020 over holiday 2019 (because of the success of the switch releasing in March and to give them more time to develop ps5)

 

[Razgriz-Specter]

Will be interesting, most likely scenario for now is Sony getting at least a one year head start, which on the other hand gives MS the opportunity to come up with a more powerful one which they could (and most likely will) announce before PS5 launches.

If PS5 releases in 2019 and plays out similar to PS4 thats likely too soon for a NextBox announcement.
It would also likely destroy the X's 3rd holiday season.

 

[FinalAres]

MS will have an issue competing with PS5, regardless of hardware, unless they do something about their first party exclusives portfolio.

I predict greedy MS will double down on power, but they'll market their machine as more premium and charge too much for it.

I predict arrogant Sony will deliver hardware that is both weaker and cheaper than the Xbox but delivers no better value for money. They'll focus on exclusives but those will take a while to come through.

I predict that Switch will be extremely popular in those first couple of years of the next generation as people aren't yet convinced to move over to the new generation of sony and microsoft consoles.

Then 2 years after the next gen starts, people will start to move over quite quickly and sales will increase. Nintendo will release a console with three screens, controlled hands-free and there still won't be a virtual console on Switch,

 

[Bitch Pudding]

If PS5 releases in 2019 and plays out similar to PS4 thats likely too soon for a NextBox announcement.
It would also likely destroy the X's 3rd holiday season.

Well, I suppose once PS5 hits the market XOX' "so powerful almost unfair" sales pitch is dead in the water anyhow. Announcing the Nextbox 1-2 years in advance to make XBOX fans (and enthusiasts) wait for their new console is just icing on the cake on the XOX's grave, unless they are able to offer it for lets say 300 bucks by then.

 

[TheThreadsThatBindUs]

Will be interesting, most likely scenario for now is Sony getting at least a one year head start, which on the other hand gives MS the opportunity to come up with a more powerful one which they could (and most likely will) announce before PS5 launches.

There's no real reason to believe that both MS and Sony won't be pushing hard to release in the same year; i.e. as soon as the relevant manufacturing technology (7nm) is ready and mature enough.

PS3 proves that launching a full next-gen console a year late is an unmitigated disaster. By allowing your direct competitor a full year uncontested in the marketplace you give up a 10-16m unit head-start in console sales, a full year head-start in entrenching consumer mindshare as well as a full year head-start of next-gen game releases, which all become effective defacto exclusives because there is not competitor box available to play on.

Once you're competitor is already 10-16m units entrenched in the marketplace, the battle for the gen is pretty much already over.

If PS5 releases in 2019 and plays out similar to PS4 thats likely too soon for a NextBox announcement.
It would also likely destroy the X's 3rd holiday season.

Even a PS5 launch in 2019 would significantly impact XB1X's 3rd holiday season (as well as PS4/Pro and XB1 sales). At that point, what's the point of buying an expensive mid-gen upgrade to a last-gen console at a $100 over a regular XB1, when you can buy a PS5 that is the new hotness that every gamer, developer and media outfit will be focused on.

I predict greedy MS will double down on power, but they'll market their machine as more premium and charge too much for it.

I predict arrogant Sony will deliver hardware that is both weaker and cheaper than the Xbox but delivers no better value for money. They'll focus on exclusives but those will take a while to come through.

The only way I see MS or Sony pushing out a box that is significantly stronger than their competitor this next cycle, is if one wins the lottery by signing with Intel or Nvidia for their CPU or GPU.

With both using AMD, it's likely that their APUs will be extremely similar, with the main differences potentially comprising memory sub-system design (Sony may go with HBCC and a lump of GPU accessible flash memory, but with a smaller GDDRx5/6 based RAM pool, whereas MS might go with a bigger GDDR6 RAM pool or even go HBM).

With both trying to avoid giving the other a full year head-start, both using the same IHV for their APU design, both using the same manufacturing process and both shooting to hit a holiday season launch in either 2019/2020 (dependent on 7nm maturity/availability), there's little reason to believe we'll see a similar 40-50% gulf in performance between next-gen consoles that we saw this gen.. (i.e. MS isn't trying to shoehorn a Kinect equivalent into their console APU design this time).

 

[STEaMkb]

nm

 

[TheThreadsThatBindUs]

And PS4/XB1 proves that launching the same year and having lower specs than your competitor can also be a recipe for disaster.

I think your conclusion here somewhat ignores the elephant in the room, i.e. the XB1 had lower specs, but also launched at a $100 premium and also came off the back of one of the worst PR debacles since the PS3's $599.

Had the XB1 with it's existing specs launched $100 cheaper than the PS4, and without all the "TV TV TV" BS and MS's always online clusterfuck, I'm pretty sure that things (at least in the US) would have turned out very different; as MS was coming into this gen. with soooo much more mindshare and goodwill in the minds of gamers and the gaming media alike.

Not only that, XB1 and PS4's initial sales trajectories started out almost neck-and-neck in the largest western territories, US and UK. It was beyond year 2 that XB1 sales started to level off.

Launching in the same year is critical in order to be competitive in the first place. After that, however, you need the games to sustain that momentum; something that this gen. MS just haven't had (due to an over-reliance on their core IP, Halo, Gears, Forza).

That's why a new paradigm of releasing mid-generation upgrades might be a good solution for all parties, because it allows them to keep leapfrogging one another, to the point where worrying about which system has the better specs becomes less of a sticking point, and the focus shifts instead to software and services. The most important element moving forward under this new model is continuity - if there's no cohesion between hardware iterations then you're seriously at risk of losing mindshare.

I'm not even sure that mid-gen upgrades as a concept will continue into next-gen. Without a strong, easily communicable marketing bullet point such as "4k" to drive it, i'm not sure it even makes any sense... it's certainly not a given.

Either way, if either Sony or MS don't establish their base platform (PS5/XBN) sufficiently enough in the first place, then a mid-gen upgrade 3+yrs in won't make a lick of difference to anyone.

Also, the most important thing in driving consumer mindshare is games; especially in the long term. MS has a lot of ground to make up here and I'm not confident they have the leadership, corporate backing nor appetite to really make a hard push at this.

It won't matter. A new Xbox releasing 1 year after PlayStation 5 would not have nearly the same power advantage as the Xbox One X currently has over the Pro.

Yup... unless they're bonkers enough to design a gigantic APU that they'll struggle to cool and will have to sell at a $500-600 price point or risk tanking their games division (aka Sony circa 2006).

 

[Bitch Pudding]

It won't matter. A new Xbox releasing 1 year after PlayStation 5 would not have nearly the same power advantage as the Xbox One X currently has over the Pro.

It doesn't have to be that (40%) faster to claim offering the "most powerful console!"

And if they want it to be significantly more powerful they could offer it for a higher price. Which is actually okay if every penny is invested into hardware, not some stupid gimmicks.

Thing is, they are just about to release a comparably expensive console by the end of 2017. I cannot see them telling their audience just one year later that they are about to launch the next one in 2019.

 

[TheThreadsThatBindUs]

If they're launching in 2019, they'll announce in 2019. It's always in the same year. Any earlier would be daft.

 

[Panajev2001a]

There's no real reason to believe that both MS and Sony won't be pushing hard to release in the same year; i.e. as soon as the relevant manufacturing technology (7nm) is ready and mature enough.

PS3 proves that launching a full next-gen console a year late is an unmitigated disaster. By allowing your direct competitor a full year uncontested in the marketplace you give up a 10-16m unit head-start in console sales, a full year head-start in entrenching consumer mindshare as well as a full year head-start of next-gen game releases, which all become effective defacto exclusives because there is not competitor box available to play on.

Once you're competitor is already 10-16m units entrenched in the marketplace, the battle for the gen is pretty much already over.



Even a PS5 launch in 2019 would significantly impact XB1X's 3rd holiday season (as well as PS4/Pro and XB1 sales). At that point, what's the point of buying an expensive mid-gen upgrade to a last-gen console at a $100 over a regular XB1, when you can buy a PS5 that is the new hotness that every gamer, developer and media outfit will be focused on.



The only way I see MS or Sony pushing out a box that is significantly stronger than their competitor this next cycle, is if one wins the lottery by signing with Intel or Nvidia for their CPU or GPU.

With both using AMD, it's likely that their APUs will be extremely similar, with the main differences potentially comprising memory sub-system design (Sony may go with HBCC and a lump of GPU accessible flash memory, but with a smaller GDDRx5/6 based RAM pool, whereas MS might go with a bigger GDDR6 RAM pool or even go HBM).

With both trying to avoid giving the other a full year head-start, both using the same IHV for their APU design, both using the same manufacturing process and both shooting to hit a holiday season launch in either 2019/2020 (dependent on 7nm maturity/availability), there's little reason to believe we'll see a similar 40-50% gulf in performance between next-gen consoles that we saw this gen.. (i.e. MS isn't trying to shoehorn a Kinect equivalent into their console APU design this time).

Unmitigated disaster when PS3 still ended up beating the Xbox 360?

 

[AmyS]

Since Xbox One X GPU is about 4.5 times more powerful than Xbox One GPU, I can't see Microsoft making a future Xbox with a GPU that isn't at least the same difference in power from XB1X GPU as XB1X's is from that of original Xbox One, even if that means not launching the next Xbox until fall 2021 (4 years, again). Microsoft could then certainly make use of the better density and performance of 7nm+

I think Sony will go for around 3X the GPU performance of XB1X and not launch PS5 before fall 2020.

 

[Bitch Pudding]

Since Xbox One X GPU is about 4.5 times more powerful than Xbox One GPU, I can't see Microsoft making a future Xbox with a GPU that isn't at least the same difference in power from XB1X GPU as XB1X's is from that of original Xbox One, even if that means not launching the next Xbox until fall 2021 (4 years, again). Microsoft could then certainly make use of the better density and performance of 7nm+

I think Sony will go for around 3X the GPU performance of XB1X and not launch PS5 before fall 2020.

So, first of all I'm very sorry for loss.

I do not think this "factor-theory" is valid / in line with Sony's approach when defining PS5's specs. They have certain targets (e.g. w/ regard to resolution, framerates, VR support), they got/get feedback from devs and they'll build that next system accordingly.

This could mean that they introduce a much faster CPU and memory solution but just a slightly more powerful GPU (cp. to XOX).

 

[Goalus]

Since Xbox One X GPU is about 4.5 times more powerful than Xbox One GPU, I can't see Microsoft making a future Xbox with a GPU that isn't at least the same difference in power from XB1X GPU as XB1X's is from that of original Xbox One, even if that means not launching the next Xbox until fall 2021 (4 years, again). Microsoft could then certainly make use of the better density and performance of 7nm+

I think Sony will go for around 3X the GPU performance of XB1X and not launch PS5 before fall 2020.

If I were MS, I'd develop two prototypes for Xbox Two, one for a launch in 2020 and the other one for launch in 2021. Then just wait and and choose one of them depending on what Sony does. And announce the XB2 before PS5 launches while making sure it has significantly higher specs than PS5.

 

[kyser73]

I don't really want this to turn into a sales thread but as it's come up...

First mover advantage in sales isn't a given. While Sony lost considerable market share reS2 in the move to 7th gen, globally it outpaced 360 sales for almost the duration of the generation and according to Adam Boyes finished ahead of the 360.

MS' issue isn't power, or features or price. Its problem is that it's image has regressed to being 'USA-box'. X1X won't address this, and unless Sony fuck PS5 up so egregiously that even the people who stayed loyal for gen 7 give it up at best MS can hope for is to get back to where they were with the 360.

Anyway...tech speculation is a lot more fun...

 

[mrklaw]

Since Xbox One X GPU is about 4.5 times more powerful than Xbox One GPU, I can't see Microsoft making a future Xbox with a GPU that isn't at least the same difference in power from XB1X GPU as XB1X's is from that of original Xbox One, even if that means not launching the next Xbox until fall 2021 (4 years, again). Microsoft could then certainly make use of the better density and performance of 7nm+

I think Sony will go for around 3X the GPU performance of XB1X and not launch PS5 before fall 2020.

If a process shrink is just around the corner then that could influence both companies. They may either choose to launch before that (with the shrink helping with a mid gen refresh, letting it take a year or so to settle down), or delay launch for a year and start the gen o the best foot possible.

 

[Shin]

They won't delay a launch for a minor node shrink, they didn't with PS4 and I don't think they will with PS5 either.
If it's available after the console has launched they'll use it for a slim model or whatever (PS5 Pro?).

 

[onQ123]

I think next time around Xbox will have the advantage of a generation-less user base so they can make the hardware & not worry too much about what software is going to sell that hardware because it's all going to be Xbox One & UWP software . But Sony will have the advantage of being able to come with something new while Xbox will most likely stay on the same path.

 

[Inuhanyou]

if sony goes full force into "next generation", MS will be forced to respond. MS doesnt want to be the only one holding their machines back after all when that's how you get eyes on your shiny new machine

IE, i think a lot of MS's "we're done with generations" talk is hot air drummed up to get people talking about their iterative unit they have going on sale

 

[Taggen86]

Just saw the DF review and Vega 64 really is a massive failure. How can a 47 percent increase in core clocks and a 40 percent increase in transistors over fury x only result in a 25 percent uplift in performance in most games? Given how inefficient their architecture is, one gets very worried about the next gen consoles and their performance leap over current gen. PS5 and Xbox two will most likely use the follow up to RX480 on a 7nm process (navi) which should result in largely the same flops and performance as vega 56 or vega 64, not very impressive considering what we could expect from next gen GTX 1160 or 1260 (volta or beyond). I really hope AMD switch architecture before Navi. Otherwise, AMD's law of diminishing returns will destroy PS5 performance.

 

[The Pope]

I predict it will release holiday 2020.
Difference between a two teraflop machine and a eight teraflop machine just isn't big enough.

 

[DonMigs85]

Just saw the DF review and Vega 64 really is a massive failure. How can a 47 percent increase in core clocks and a 40 percent increase in transistors over fury x only result in a 25 percent uplift in performance in most games? Given how inefficient their architecture is, one gets very worried about the next gen consoles and their performance leap over current gen. PS5 and Xbox two will most likely use the follow up to RX480 on a 7nm process (navi) which should result in largely the same flops and performance as vega 56 or vega 64, not very impressive considering what we could expect from next gen GTX 1160 or 1260 (volta or beyond). I really hope AMD switch architecture before Navi. Otherwise, AMD's law of diminishing returns will destroy PS5 performance.

Yeah considering it's even supposed to have tiling support it is a bit of a letdown. Something around GTX 1080 level isn't enough of an upgrade over PS4 Pro.

 

[napata]

Yeah considering it's even supposed to have tiling support it is a bit of a letdown. Something around GTX 1080 level isn't enough of an upgrade over PS4 Pro.

Some of you guys really have unreasonable expectations. A GTX 1080 is more than 2.5 times as strong as a Pro. You might not even get something on par with a 1080.

AMD hasn't been able to scale down a Fury X to a console tier GPU in 2.5 years. What makes you think they'll be able to scale down a VEGA 64, which uses even more power than a Fury X btw, in 2.5 years. Technology is slowing down, not speeding up.

 

[pelican]

If they're launching in 2019, they'll announce in 2019. It's always in the same year. Any earlier would be daft.

I agree. Probably similar timings to the PS4 reveal. Unless they want to show some transparency similar to MS with the X. But, that is doubtful considering previous behaviour e.g. radio silence about the slim despite it being visible in a UK retailer/ digital foundry video.

 

[Ushay]

I think next time around Xbox will have the advantage of a generation-less user base so they can make the hardware & not worry too much about what software is going to sell that hardware because it's all going to be Xbox One & UWP software . But Sony will have the advantage of being able to come with something new while Xbox will most likely stay on the same path.

I think ultimately the publishers will have the final say, the platform owners do after all have extensive talks with them before 'jumping' generations.

Personally I can see the generation thing phasing out for a lowest common denominator model like phones, but with a wider gap between 'minimum spec'.

 

[DonMigs85]

Some of you guys really have unreasonable expectations. A GTX 1080 is more than 2.5 times as strong as a Pro. You might not even get something on par with a 1080.

AMD hasn't been able to scale down a Fury X to a console tier GPU in 2.5 years. What makes you think they'll be able to scale down a VEGA 64, which uses even more power than a Fury X btw, in 2.5 years. Technology is slowing down, not speeding up.

Ugh, hopefully they delay as much as possible. PS5 ending up with just GTX 1070-level performance with Vega 56 isn't much of an upgrade over the RX 470-level Pro.

 

[onQ123]

I think people are too caught up on the number of GPU flops when next gen will most likely enable effects that we can't get so easy on PS4/Xbox One , the smart thing to do is offload things to special hardware that will do it more efficiently much like the ID buffer hardware , I'm sure if that was something you had to do with the normal GPU it would take up a lot of compute time & memory bandwidth.

Remember PS4 Pro wasn't suppose to be able to play any 4K games according to NeoGaf.

 

[c0de]

Remember PS4 Pro wasn't suppose to be able to play any 4K games according to NeoGaf.

Point us to people who said that in masses. You always have some doubters but this was never general consensus. People like me even argued for native 4k games running on PS4 amateur and Xbox One.
These generalizations don't help at all in any discussion.

 

[onQ123]

Point us to people who said that in masses. You always have some doubters but this was never general consensus. People like me even argued for native 4k games running on PS4 amateur and Xbox One.
These generalizations don't help at all in any discussion.

It was pages of back & forth whenever someone said anything about 4K games on PS4 Pro before it was released but that wasn't the point of my comment , I'm just saying that when people here looked at the basic specs they didn't get the full picture & it will be the same way if we get PS5 specs & everyone see something like 10TFLOPS they will think it's only a little bit more powerful than PS4 Pro & Xbox One X but you can't just go by tflops without looking at the other advances.

 

[TheThreadsThatBindUs]

Unmitigated disaster when PS3 still ended up beating the Xbox 360?

Until Sony managed to put out the slim PS3 it was a complete disaster. Sony did an excellent job of pulling it back in the end, but initially, PS3 was a mess.

 

[TheThreadsThatBindUs]

I don't really want this to turn into a sales thread but as it's come up...

First mover advantage in sales isn't a given. While Sony lost considerable market share reS2 in the move to 7th gen, globally it outpaced 360 sales for almost the duration of the generation and according to Adam Boyes finished ahead of the 360.

MS' issue isn't power, or features or price. Its problem is that it's image has regressed to being 'USA-box'. X1X won't address this, and unless Sony fuck PS5 up so egregiously that even the people who stayed loyal for gen 7 give it up at best MS can hope for is to get back to where they were with the 360.

Anyway...tech speculation is a lot more fun...

There's first mover advantage, and there's giving your direct competitor a full year alone in the marketplace uncontested. Regardless of the latter success of the PS3 (owed to Sony's efforts in price reduction of the PS3 as well as their wider reach in international markets - i.e. a luxury MS cannot boast), Sony still learned a valuable lesson about giving your competitor too much time alone in the market.

It's not about who goes first, so much as how much time they have in the market to generate a lead and soak up mindshare.

 

[TheThreadsThatBindUs]

They won't delay a launch for a minor node shrink, they didn't with PS4 and I don't think they will with PS5 either.
If it's available after the console has launched they'll use it for a slim model or whatever (PS5 Pro?).

if sony goes full force into "next generation", MS will be forced to respond. MS doesnt want to be the only one holding their machines back after all when that's how you get eyes on your shiny new machine

IE, i think a lot of MS's "we're done with generations" talk is hot air drummed up to get people talking about their iterative unit they have going on sale

Agreed

 

[Inuhanyou]

Just saw the DF review and Vega 64 really is a massive failure. How can a 47 percent increase in core clocks and a 40 percent increase in transistors over fury x only result in a 25 percent uplift in performance in most games? Given how inefficient their architecture is, one gets very worried about the next gen consoles and their performance leap over current gen. PS5 and Xbox two will most likely use the follow up to RX480 on a 7nm process (navi) which should result in largely the same flops and performance as vega 56 or vega 64, not very impressive considering what we could expect from next gen GTX 1160 or 1260 (volta or beyond). I really hope AMD switch architecture before Navi. Otherwise, AMD's law of diminishing returns will destroy PS5 performance.

To be real for a second, even if PS5's GPu had the equivalent flops of Vega 54, that's still 10tflops, which is already in line with what we expected for a next gen PS5.

 

[Genio88]

Makes sense, that seems to be how console will evolve from now on, a bit like iPhones/iPad...i hope that Nintendo will do that with Switch too, Switch is the only console i care about, besides PC, and i'd like a Switch Pro in 2/3 years with backward and forward compatibility

 

[kyser73]

Until Sony managed to put out the slim PS3 it was a complete disaster. Sony did an excellent job of pulling it back in the end, but initially, PS3 was a mess.

There's first mover advantage, and there's giving your direct competitor a full year alone in the marketplace uncontested. Regardless of the latter success of the PS3 (owed to Sony's efforts in price reduction of the PS3 as well as their wider reach in international markets - i.e. a luxury MS cannot boast), Sony still learned a valuable lesson about giving your competitor too much time alone in the market.

It's not about who goes first, so much as how much time they have in the market to generate a lead and soak up mindshare.

Globally the PS3 outsold the 360 from launch - the only months it didn't were around the Kinect launch period IIRC.

It never 'came back' except for the US market, and the whole 'PS3 comeback' narrative needs to die in a fire.

 

[hydragonwarrior]

4k/60 fps