-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
PS5 and next Xbox launch speculation - Post E3 2018
- Thread starter Phoenix Splash
- Start date
MLB The Show is just not an annual sports title to Sony. It's AAA as much as any of their other blockbuster titles.
Canada local? If it's local doesn't mean much other than that they release new hardware all the time especially with accessories in the mix. I do believe it's a vacant position since it's a contractor.
But we'll see how it goes I believe until the end of the year we'll have some news even if it's only dev kits.
Don't be so sensitive, dude.
The issue is that these thread often get so cyclical when perfectly reasonable speculation based on a knowledge of the industry, gets called out by some new entrant into the thread who's basing their take on PC retail components. Then folks have to explain why for the next few pages... rinse and repeat.
Also, given that there are no 7nm CPU or GPU currently available on the market, equally the person's speculated specs in question provided no indication of ALU count on the GPU, i'm not sure what you're basing your die size claim on at all.
We have no idea what transistor densitites 7nm will deliver, nor what the Navi microarchetecture looks like, nor the exact config of the PS5, nor even how much of the die will be devoted to i/o, cache or the memory interface. So any speculation on die size for a given TFLOPs perf number is literally finger in the air, pie in the sky. Not to mention that TFLOPs performance is based on GPU clockspeed, which for a mainline console isn't going to be clocked as aggressively as the Xbox One X, so any comparison there won't be very useful.
In the end, the die size of a PS5 may in fact be larger than the XB1X, but not by much, and it doesn't need to be to deliver 10-13TFLOPs of SP GPU compute -- you're only talking 80 CUs at 1270 MHz, which should be more than doable on 7nm for < 370 mm2 (i would imagine).
MLB the show was already mentioned. I don't think that counts as the major unannounced title in the context of the discussion because we all know that is coming, so it's as if it's announced.
Benji is not playing tricks with us
!
Good catch. So many new positions...2019 sounds more and more probable.
I think Sony will surprise many of the "people in the know" this time around, seems like they've been keeping it really quiet to avoid leaks
~+10% real world performance improvement. I mean, it's something alright. Could be just as easily reached by clocking optimization though which wouldn't require s/w support or h/w complexity increase.
72, not 80. You will never get full GPU sp activated.
If there is one it's probably been moved to PS5 by this point. I don't see the point to announce a new big game for PS4 in 2019. It should have been this year and Shawn Layden is saying they don't have much for 2019.
I said to myself they are going to announce last games for PS4 in 2018. I will not go through that again. I was very disappointed in 2018.
Sports games tend to really, really like being cross-generation. One of the sports games was the last to release for PS2 before console support died. Sports games are those that I don't expect to become PS5 exclusives, but be PS4 games with PS5 upgrades.
I'm assuming 80 activated. I have no idea what the SE layout looks like, so no idea whether 88CUs is realistic, but that was my assumption.
The problem with moving a game from the PS4 to the PS5 is the install base. It's the difference from selling millions and millions to a few hundred thousands.
So it could be done on some cases, specially if it's a smaller project, but it can't be done on most.
Next generation will not be boring from a technological point of view if one or the two consoles have this rasterizer technology described by this AMD patent.
Parallel micropolygon rasterizers
https://www.freshpatents.com/-dt20180301ptan20180061124.php
This rasterizer with Mesh shader/primitive shader rendering open crazy possibility...
Mesh shader is a must, a sucker punch guy is excited by the possibility of mesh shader:
http://reedbeta.com/blog/mesh-shader-possibilities/
We can do adaptative tesselation and made tessellation really useful next generation, it is possible with compute shader but it goes faster with Mesh Shader 25% faster without optimization. We can use tessellation for other things than terrain rendering , Catmull Clark subdivison or other subdivison algorithm and displacement mapping if we have Mesh shader/primitve shader and this parrallel micropolygon rasterizer inside next gen GPU... Replace sometimes normal maps or parallax mapping by true geometry...
See all the possibility and technical improvement but it helps content production too and make animation easier:
https://developer.amd.com/wordpress/media/2012/10/Tatarchuk-Tessellation_GDC08.pdf
and funny things:
Parallel micropolygon rasterizers
https://www.freshpatents.com/-dt20180301ptan20180061124.php
This rasterizer with Mesh shader/primitive shader rendering open crazy possibility...
Mesh shader is a must, a sucker punch guy is excited by the possibility of mesh shader:
http://reedbeta.com/blog/mesh-shader-possibilities/
We can do adaptative tesselation and made tessellation really useful next generation, it is possible with compute shader but it goes faster with Mesh Shader 25% faster without optimization. We can use tessellation for other things than terrain rendering , Catmull Clark subdivison or other subdivison algorithm and displacement mapping if we have Mesh shader/primitve shader and this parrallel micropolygon rasterizer inside next gen GPU... Replace sometimes normal maps or parallax mapping by true geometry...
See all the possibility and technical improvement but it helps content production too and make animation easier:
https://developer.amd.com/wordpress/media/2012/10/Tatarchuk-Tessellation_GDC08.pdf
and funny things:
Last edited:
Agreed. Some people are getting too invested into this that they lash out in subtle ways when speculation changes due to updated info or more digging. I doubt many remember everything they posted 6 months ago, so to hold him accountable when he's proven to be genuine with his posts is just sad. I hope he doesn't come back to this thread. Maybe then people will think twice before feeling the need to be a dick over something like launch year speculation.
This is the type of stuff that excites me about next gen. Hopefully both next gen consoles will feature this and help adoption.
It's a performance optimization of the same old rasterisation process, why would it mean that next gen won't be boring? Something like "Mesh shaders" are possible on consoles since PS2 era and it's just an optimization of geometry processing as well. Do more triangles and faster rasterisation sound exciting to you?
Or do what MS did with Tital Fall, market only PS5 version and forget the PS4 version even exists.
A micropolygon friendly GPU is exciting for me... Currently subpixel polygon are not GPU friendly... It opens tons of possiblity REYES style motion blur and depth of field(Defocus blur) and plenty of other things...
EDIT: like in this 2009/2010 research paper
http://graphics.stanford.edu/papers/hwrast/
EDIT2: And good tesselation with adaptative tesselation from day one.... ;)
Edit3:
PS2 did not offer all the possibility of Mesh shader not enough power it was just more flexible than vertex/geometry/domain/hull shader.
If it was the base of the architecture geometry pipeline would not have been a mess post PS2...
Last edited:
Oh wow. Limited term contract position, no longer than 12-15 months sounds pretty definitive.
Indeed that there's tons of possibilities with this I'm really really excited
Current GPU aren't micropolyon friendly... Micropolygon are described as subpixel polygon... it is a little more than more polygons, this is why I insist about Mesh shader/Primitive shader coupled with parrallel micropolygon rasterizer.. Going further than Turing the best rasterizer GPU architecture...
The only missing feature is available on AMD GPU use in Dreams(SIGGRAPH 2015 presentation) but I am not sure it is exposed in DirectX
EDIT: most of the things inside this GDC 2008 presentation were not realised because the geometry pipeline was a mess
https://developer.amd.com/wordpress/media/2012/10/Tatarchuk-Tessellation_GDC08.pdf
Last edited:
Micropolygon rasterisation has nothing to do with actual Reyes micropolygonal rendering. It's also highly likely to be even less efficient than straight raytracing for real time uses. I'm also pretty sure that all modern GPUs have stuff which helps them with below-quad triangle rasterisation already.
I should clarify that the "yes" pertains to whether or not I'm excited about these things.
This rasterizer hardware created for a Stanford research looks like a lot the rasterizer described by the AMD patent. It looks a bit less advanced because it can't work with big triangle like the rasterizer described in AMD patent. But you can use high quality motion blur and high quality defocus blur...
http://graphics.stanford.edu/papers/hwrast/
And no current GPU aren't subpixel polygon friendly and it is one of the reason good occlusion culling is a must...
And I did not say REYES rendering but motion blur and defocus blur REYES style... Geometry caching REYES rendering is not a good solution for realtime solution, deffered tesselation is probably a better one...
EDIT: Another problem with REYES is shading before occlusion culling. It is done because of shading reuse and displacement shading but not very realtime friendly....
Last edited:
The only thing we can be sure about is that #TeamH22020 is wrong. It's going to be Q4 2019 imo. but it could be Q12020. I'm just sitting idly by, sharpening my pitchforks, waiting to roast some "insiders" if it's Q42019. If I never hear the word insider again, it'll be too soon.
No I understood and one last things it is offline rendering but micropolygon or subdivision surface are compatible with raytracing. It is offline rendering but subdivd surface with direct raytracing:
https://graphics.stanford.edu/~boulos/papers/subdiv_tr.pdf
Hybrid rendering FTW
Agreed. H2 2020 is not happening. If it's not Q4 2019, it's Q1 2020.
Anyone who thinks it's Q4 2020 needs to hand me whatever they smoking.
Oh shit, that marketing job posting is legit. I
Well I think it's worth thinking about what this timeline means in terms of marketing. Surely the campaign for PS5 would start before the console launches.
Edit: I went back and realized you folks are now talking about that logistics position. Still seems just as likely that it could mean Q1 2020 to me. They're tasked with getting the pieces into place up to the launch, but it seems like they're done after that.
Edit: I went back and realized you folks are now talking about that logistics position. Still seems just as likely that it could mean Q1 2020 to me. They're tasked with getting the pieces into place up to the launch, but it seems like they're done after that.
64% of the people still believe it and this logistics position means 2019 IMO
Each day we keep switching between 2019 and 2020...!
I think you make a very good point here in that these jobs were posted just a few days after Kenichiro Yoshida spoke with the FT which some of us noted was pretty significant for Sony. Surely not a coincidence.
DriftingSpirit Keep calm and breath....;)
We'll also have some additional news regarding dev kits most likely by the end of the year.. Especially since Benji said that they already using Pre Built PCs as temporary dev kits.. Most likely dev kits will be finished by this fall and be on studios in q1 2019.
Also for those who are curious whether it's 2019 or 2020 just wait till March by then if there's no PS Meeting most likely guarantee is that it is indeed 2020.
Good choice. Q4 2020 team better start switching fast because ya'll gonna look silly soon.
Yeah I hope something comes off that is more directly PS5 related like a confirmed dev kit. I think more would accept such more readily than the crumbs we've had so far. I agree about March being the absolute latest cut-off point for definitive info.
Change your vote now. You'll thank me down the line.
#Q42020isfakenews
Imagine the possibilities with the Ryzen CPU when it comes down to Open World games and client side calculations with the console optimization...
If we're taking 12-16 months at face value, it could be holidays 2019 or spring 2020
I don't know what to believe anymore. I was sure it would be Q4 2020, but sings are now pointing to Q4 2019 or Q1 2020.
I honestly now don't care when it comes, just as long as the PS5 supports physical media and has BC with PS4 games. Day one if they can nail that.
I honestly now don't care when it comes, just as long as the PS5 supports physical media and has BC with PS4 games. Day one if they can nail that.
Does AMD have everything as well? Of course it remains to be seen how comparable the primitive vs. mesh shader implementations are.
Of course, it's easy to forget PS4 games aren't even taking advantage of FP16 due to it only being in Pro. Lots of things adding up to show that a Navi TF does not equal a Polaris TF.
Also, team 2019 be like:
Does AMD have everything as well? Of course it remains to be seen how comparable the primitive vs. mesh shader implementations are.
Of course, it's easy to forget PS4 games aren't even taking advantage of FP16 due to it only being in Pro. Lots of things adding up to show that a Navi TF does not equal a Polaris TF.
Also, team 2019 be like:
No one out of AMD know exactly what primitive shader is doing but after Mesh shader announcement, I saw multiple dev tweet on twitter thinking it is like primitive shader if it is not the case I hope we will have it inside primitive shader 2...
OP
OP
There's so much new info and I do not think that I understand all of it, but it seems like it's legit information coming from Sony themselves. Did any of you guys find if there was a similar job offer to this one in 2011-2012(1-2 years before the PS4)? Maybe that could give us even a better idea about their roadmap.
I think I'll be adding this new info to the OP.
I think I'll be adding this new info to the OP.
So (not a dig at but) did Benji try some damage limitation when the 2019 gang started building up steam of things he was saying?
Or are a lot of people here really going to have a big bubble burst!?
lol, no chance 'just wait 5 months' ;)
Or are a lot of people here really going to have a big bubble burst!?
lol, no chance 'just wait 5 months' ;)
I'm still trying to determine if the rumors of Sony being so hands on with Navi is a good thing or a bad thing in this case. If Sony is more on the pulse of needed features than AMD, it's not a great look for AMD. Similarly, if Sony is behind the curve, they're hampering a whole generation of console designs.