Maybe , but i think the game is more bandwith bound rather than ALU bound.
Digital Foundry Article Technical Discussion Archive [2014]
- Thread starter DieH@rd
- Start date
-
- Tags
- digital foundry
- Status
Maybe , but i think the game is more bandwith bound rather than ALU bound.
Or there may be other render targets they preferred to be kept around in there as well.
It'd fill the ESRAM with no room for anything else! ESRAM needs to be used as working space as well. 900p would use 22 MB, leaving 10 MB for other data if it's stored complete in ESRAM.
Looks like it might actually be a 980SLI. Assuming the test is comparable to the CPU scaling test then you're looking at 69fps average on the the FX4300 and 81fps on the FX8350.
http://gamegpu.ru/action-/-fps-/-tps/lords-of-the-fallen-test-gpu.html
I'd like to now if there's an input latency hit on the dynamic resolution. If they are scaling the res with no penalties, you have to wonder why not do that on every platform? But given resolution and framerate are locked in MP, I'm guessing there's an extra bit of lag in the single player to support dynamic res, which they didn't want to include in the MP experience (which would be ironic given net latencies!).
There might be a small hit but that hit has got to be less than losing frames right?
I think the reason it exists in SP is because the dynamic resolution isn't fully dynamic they were able to pinpoint where and when frames were going to be loss. In multiplayer they don't have such control so maybe they removed it all together?
Or
Option 2
It was a choice made because they wanted to remove any possibility of variation in competitive multiplayer. Which I agree with. You don't want anything to change in that environment. Otherwise players are likely to blame the game than themselves for not being successful.
I think adaptive v-sync is allowing more breathing room than engaged v-sync. Why not allow PS4 users the option of which v-sync to choose?
Last edited by a moderator:
Really odd choices for PS4. Why wouldn't they put adaptive v-sync and dynamic resolution in it? Assuming it has headroom, it probably wouldn't drop resolution anyway, and v-sync would help it stay at pretty much 60Hz. Seems odd. The way it is, I'd probably want to play campaign on X1 and multiplayer on PS4. My guess is they'll have some kind of a patch to improve framerate on PS4, probably by changing v-sync.
D
Deleted member 11852
Guest
It's like they had different priorities for each platform, 60fps on Xbox One and 1080p on PlayStation 4. Weird.
hesido
Regular
You are much more tech savvy then I am, but I need to add that fighting with local latency is a different ball game than net latency. The user can see instant reactions from his own input locally (aiming, shooting) which is just fine, and how latent or early those actions are met in the net-side is a different aspect. So fighting local latency has instant benefits whether you are online or not.
I think the results are because they considered the performance on PS4 sufficient earlier on, and then spent more time solving the performance issues on the Xbox. They did a good job with that, and got that finished just in time, but not in time to then consider taking that back into the PS4 version. May still happen at some point though.
I've updated the results in techingames.net btw.
I'd almost ask instead: why not do that in every game ever that you're not 100% certain can keep 60fps? Who knows we'll find out.
Not really, considering how decoupled local control latency is from network latency in most MP FPS games.
Advanced Warfare on Xbone feels like it has dynamic resolution so that it can go up to 1080P when there is nothing on screen rather than dropping from 1080P when there is a lot on screen so as to save framerate.
The article says that anytime you are in combat it basically appears drops down to the lowest dynamic resolution - but they want to do some additional measurements to rule-out if it doesn't use intermediate resolutions some or a lot of the time.
And risk accusations of p p p parity?
Actually, could some of the drops on ps4 could be cpu related? Latest Ubi slides showed X1 CPU being about 15% faster. Especially if you drop X1 res to fall back into esram and ease main memory contention, there could be a performance delta starting to open a little there.
djskribbles
Legend
Depends who you ask. Native 1080p with arguably small drops in sp and stable framerate in mp, vs variable res.in SP, sub 1080p mp with some tearing. Pick your poison.
I've yet to see anyone complain about fhe framerate on ps4.
- Status
Similar threads
