Based on the fact they're advertising it as RDNA2 part, yes, the hardware should be capable of it. Until proven otherwise, you can assume feature parity between the consoles on graphics side of things (API limitations aside)
Current Generation Hardware Speculation with a Technical Spin [post GDC 2020] [XBSX, PS5]
- Thread starter Proelite
- Start date
- Status
It's most likely standard RDNA2 features.
VRS is basically a tweaked MSAA.
It certainly is not the most important features after RT.
I'm not even sure if RT is the most important feature.
Getting rid of old vertex shader pipeline is quite a big one, developers can now write nice tessellator and lot's of interesting things with Mesh Sahders.
New texture samplers and other additions are also quite big ones, we will see a lot of interesting takes on shading, shader reuse and so on. (Perhaps quite silly things, like during runtime creating narrow/cone area lightfields with RT for far reflections stored in textures.. etc.)
Pretty sure it's RDNA2 features, so yes I'm certain of it.
API limitations have finally started to open even on PC side, will be very interesting to see if there is still some on DX12/Vulkan.
It was interesting to see mentions of barycentric coordinates being accessed on Ps4, when GCN/APIs didn't allow it on PC.
Silent_Buddha
Legend
It's 0.1 GHz when comparing XBSX CPU with SMT versus PS5 CPU unknown.
It's 0.3 GHz when comparing XBSX CPU without SMT versus PS5 CPU unknown.
If the PS5 CPU is with SMT at 3.5 GHz, it's going to consume more power than PS5 CPU without SMT which in turn means there's a far larger chance that the PS5 will have to downclock the GPU.
Of course if a developer is pushing the PS5 GPU particularly hard, then the PS5 CPU may not be able to hit 3.5 GHz regardless of SMT or no SMT.
That said, I don't expect this to lead to any large disparity in presentation between the two consoles.
The only practical difference this will make, IMO, is that developers on PS5 have to pay attention to how they use the CPU and GPU where on XBSX they get the same performance from each regardless of how they use it.
Regards,
SB
I'm not saying XBSX has a meaningful advantage! It's a qualitative difference between the advantage of SSD speed versus CPU speed, and not a quantitative advantage that one is much more better off. A CPU speed advantage will affect every game on the platform with no work needed by the devs, whether that's by 10% or 0.1%. Have a game that's maxxing out the system? Then a faster CPU makes it run faster.The SSD speed advantage OTOH is likely going to be based on XBSX's performance for now as it's a new tech devs don't know how to use or scale ideally that's already some two orders of magnitude faster than the 20 GB/s target streaming of PS4. For a dev to make their game better on PS5 thanks to its faster SSD, they'll have to deliberately develop for it, unlike making use of XBSX's faster CPU that just happens. Maybe down the line, engines will scale naturally based on storage performance just as they do graphics performance and CPU performance (we have that now regards pop-in on streaming assets at least), but certainly for the time being, PS5's double-speed SSD isn't going to yield massive advantages in cross-platform games. Cross-platform games will target the most sensible storage performance target and not care about optimising for a niche audience.
Considering one is advertised as best case "boost" and the other is locked in, we actually don't the speed difference between the two.
The SSD speed advantage OTOH is likely going to be based on XBSX's performance for now as it's a new tech devs don't know how to use or scale ideally that's already some two orders of magnitude faster than the 20 GB/s target streaming of PS4. For a dev to make their game better on PS5 thanks to its faster SSD, they'll have to deliberately develop for it, unlike making use of XBSX's faster CPU that just happens. Maybe down the line, engines will scale naturally based on storage performance just as they do graphics performance and CPU performance (we have that now regards pop-in on streaming assets at least), but certainly for the time being, PS5's double-speed SSD isn't going to yield massive advantages in cross-platform games. Cross-platform games will target the most sensible storage performance target and not care about optimising for a niche audience.
But regardless of streaming, surely when loading the game, loading levels, fast travelling and all of that will simply load at the maximum possible speed on either?
Are you just assuming this, or has there been more info released somewhere?
anexanhume
Veteran
I came across this, so it does indeed appear to be primitive/mesh shader related:
https://fuse.wikichip.org/news/3331/radeon-rx-5700-navi-and-the-rdna-architecture/2/
I’m not sure what kind of foothold they have. I remember reading about their tech and being very intrigued, but it also sounds like Kraken has a huge foothold.
D
Deleted member 7537
Guest
The SSD speed advantage OTOH is likely going to be based on XBSX's performance for now as it's a new tech devs don't know how to use or scale ideally that's already some two orders of magnitude faster than the 20 GB/s target streaming of PS4. For a dev to make their game better on PS5 thanks to its faster SSD, they'll have to deliberately develop for it, unlike making use of XBSX's faster CPU that just happens. Maybe down the line, engines will scale naturally based on storage performance just as they do graphics performance and CPU performance (we have that now regards pop-in on streaming assets at least), but certainly for the time being, PS5's double-speed SSD isn't going to yield massive advantages in cross-platform games. Cross-platform games will target the most sensible storage performance target and not care about optimising for a niche audience.
No, I didn’t want to mean you were saying that. I just wanted to say that I really don’t expect games to hit CPU limits on these consoles anytime soon.
Edit: everyone should take a look at the latest steam survey to understand why.
I wasn’t drawing any conclusions, just stating that in terms of hardware this seems like a different balancing of virtually identical silicon ... ? And so I am interested in how the differences play out. I can certainly imagine that the advantage may be significant.
DavidGraham
Veteran
Assuming it doesn't downclock under load.
Yep.
That said, can the difference be that big?
Of course it's possible that on PC everything is just so badly optimized, but so long as you set SATA SSD as baseline, going faster SSDs and NVMe seems to have relatively small effect on loading times?
I find something about it where they tell the last 10 years I/O software system are obsolete and we deliver something new about it.
anexanhume
Veteran
This is going to get old real fast.
Like Cerny said, just a 2-3% downclock could reduce thermal by 10%. That's ~50MHz and shouldn't be much greater. so it's still pretty much a 10 TF console through and through with a sizable 350MHz clock advantage in worst case scenario while keeping a impressive 405MHz advantage most of the time.
DuckThor Evil
Legend
That's because the SSD speed isn't the bottleneck, but the low CPU utilization on decompressing the data being loaded.
That's not how the XSX memory setup works. There are 10 32-bit controllers, of which 6 attach to 2GB chips and 4 attach to 1GB chips. Every controller works at 14Gbps. The 10GB "fast pool" is just the first 1GB on all the chips, so the 6GB "slow pool" is not independent of the fast pool, having the CPU use the slow portion does still slow down the fast portion.
That's why the total bandwidth is still 560GB/s, not 560+336GB/s
The full twitter convo:
Yea DF: get the scoop and start talking
Yea DF: get the scoop and start talking
- Status
Similar threads
- Locked
