Yeah but let's not focus on third parties. I am talking about MS's first party studios. You are explaining why third parties are less likely to use them.
Do you think there will be a mid gen refresh console from Sony and Microsoft?
- Thread starter invictis
- Start date
If you look at how many AAA first-party titles have been released for Series consoles so far, it's no wonder that the games so far barely used the console technically.
Halo Infinite and FH5 cross-platform games. Starfield runs on an ancient game engine. Technically, Forza Motorsport uses a couple of great things that make it capable of Ray-tracing at 60 FPS on the console.
Still, what can be really so advanced is Hellblade 2, then Avowed, Indiana, Fable, Gears, etc. Amazing, but these "nextgen" games are just starting to come out...
Halo Infinite and FH5 cross-platform games. Starfield runs on an ancient game engine. Technically, Forza Motorsport uses a couple of great things that make it capable of Ray-tracing at 60 FPS on the console.
Still, what can be really so advanced is Hellblade 2, then Avowed, Indiana, Fable, Gears, etc. Amazing, but these "nextgen" games are just starting to come out...
That's a good point. Sometimes things sound good on-paper, but either fall short in results in reality, require too much work, or come with side effect downsides that the positives dont outweigh enough.
I feel like we've been getting far less technical discussions/interviews with developers about this stuff in the past couple years. Hard to know what's going on.
Same issues as third parties. You still need to deliver a game. You only integrate systems you need. If it’s not a high priority yet, don’t plumb it in.
Might be worthwhile to wait for HB2 and games forward from there. Games that are designed to push graphical boundaries make sense to use everything that is available.
A good example is sampler feedback streaming. It’s a great idea!
Problem is, it requires tiled resources. And no one uses tiled resources. Barely anyone does at least - the only game that comes to mind is maybe Gears 4/5.
Edit: I’ll add to the list as I Google-fu
And Halo
Probably flight simulator
Maybe a slew of UE titles on Xbox.
Last edited:
Tiled resources is terrible. Developers mostly only use it for small things like mipmap streaming or some buffer resources ...
Yup. It seems to be largely an Xbox only construct. I rarely see it used outside of Xbox consoles. I’m not sure if this will change with work graphs? I don’t understand enough about it yet to ask a question
Perhaps if MS writes its own engine for the Series consoles, it could be a solution to take advantage of these functions, like the hardware SFS. But apparently they chose the more comfortable way of UE5... Most of their internal titles run on this engine, the exception being perhaps the two Forza studios - it will be interesting to see what Fable uses.
The Coalition will probably put a better version under Gears 6 than UE5.5, but that studio has exceptionally high-level programmers.
One month and we'll see what Ninja Theory can do, first.
The Coalition will probably put a better version under Gears 6 than UE5.5, but that studio has exceptionally high-level programmers.
One month and we'll see what Ninja Theory can do, first.
Tiled resources is just an awkward way of exposing virtual memory support in modern GPUs. Quite a few Sony games like Death Stranding and Marvel's Spiderman uses the feature on PS/PC but nobody really makes any interesting use out of it for things like virtual texturing (original intention of functionality) or accelerating virtual shadow maps (eventually software raster was needed to achieve 16K res shadow maps) ...
All modern gfx APIs including every modern console gfx API supports the feature. It's probably best to put the feature to rest rather than iterate on it because neither AMD nor Nvidia have any desire to improve upon it. The best hardware implementation of tiled resources I've seen was on Intel graphics where their hardware included a fast path to update the page table mappings ...
I would say, not so much TAA, rather the other denoisers that are Not RR + TAA.
I think a lot is to do with the way MS demand that every piece of software delivers maximally for all their main goals: Gamepass, streaming, and "monetisation of engagement" or some shit like that.
Every game has to: sell on as many platforms as possible, drive Gamepass on as many Gamepass supporting platforms as possible, stream to devices where people haven't even invested in the hardware to begin with, and be as monetiseable as possible (meaning again, run on as many platforms as possible). Only the death of last gen consoles has allowed MS to move the baseline up, they didn't chose to do this to give Xbox killer apps.
Fully using Series features and making killer software helps the Series consoles, but that comes way down MS's list of priorities.
And going further and building games around a baseline of Velocity architecture, Mesh Shader, RT, and advanced upscaling is only going to limit opportunities to sell games to e.g. lower end PC gamers, or grow the Gamepass market, or monetise monetise monetise. So if you'd proposed doing that back in 2019 or 20 or 21 or 22 you'd have probably been given the boot. MS don't see it as being their job to move any faster than anyone else who makes games (and they're probably behind the likes of Rememdy).
Having a successful console was supposed to be MS's ace in the hole for building Gamepass, having a streaming service, and pushing a unified DX12U feature set, similar to how Xbone was supposed to allow MS to take over the TV. But all this comes at a cost to their core console market which they consistently deprioritise. The Golden Goose only has so many hit points: you keep cutting into it eventually it'll die.
MS focued so little on Series consoles that eventually customers decided to follow their lead. MS have withered their own console business, to the point where it's going to hurt Gamepass and also limit marketability of their cloud gaming offering.
All of which is building to this comment about a mid gen refresh Xbox:
The refreshed Series X that is likely coming this year should focus on cost reduction and modest clock bumps if they're cheap enough to get. If MS can implement boost clocks on the GPU in a cost effective manner they should do this too, becuase it will improve games while requiring no work from MS on the games side, because they won't do it. It should include no new features, because MS won't use them.
This explanation makes more sense
Johnny Awesome
Veteran
Nah. MS has overall delivered comparable games to Sony this gen and this year they will begin to pull ahead.
I think the issue is that all the devs they acquired in 2017 onward are just now getting to putting out their new games.
I'm not doom and gloom at all on Xbox as a hardware platform and I think they should go for broke.
They should shrink the X, but they should put out a new console in 2025 and another in 2029. Set the expectations that they will be a hardware leader.
All this running with their tail between their legs is loser talk.
The idea that they can't support 3 or 4 configurations is nonsense. The cross gen period proves they could.
I think the issue is that all the devs they acquired in 2017 onward are just now getting to putting out their new games.
I'm not doom and gloom at all on Xbox as a hardware platform and I think they should go for broke.
They should shrink the X, but they should put out a new console in 2025 and another in 2029. Set the expectations that they will be a hardware leader.
All this running with their tail between their legs is loser talk.
The idea that they can't support 3 or 4 configurations is nonsense. The cross gen period proves they could.
davis.anthony
Veteran
How so?
Developers have catered to varying levels of hardware features across multiple platforms for decades.
If you were to graph that, I reckon it's a very downward trend. In the past, they had to as all the hardware was different, but costs weren't insane to do so. And even then the more exotic flavours of hardware were often overlooked. Now, it's not necessary to target specific hardware variations while managing to make a game that runs okay, with the optional extras being very optional. We've seen hardware features go largely unused in the past few gens.
As the hardware has homogenised and costs to just make a game, let alone optimise it, have increased, the economics of targeting hardware specifically have changed over those decades so I don't think the past is particularly informative.
cheapchips
Veteran
Fable's using the Forza engine. It'll be interesting to see postmortems how that went. I can't help think EA's 'Frostbite for every genre' disasters and Creation Engine's hats as trams.
Bespoke Xbox/PlayStation engines aren't what's needed to take full advantage of their hardware. It's production time and resources allocated to doing so.
Specifically with regard to mid gen refreshes - the best thing Sony can do it make it really easy for any engine/developer to implement it's DLSS equivalent thingy. Sounds like they are.
For MS's part, even although they're skipping mid-gen, I'm half expecting their DLSS equivalent to be injected into BC titles with no developer intervention.
Last edited:
Quality of the titles are tangential to the goals of the console. One doesn’t have to suffer for the other to improve and vice versa.
MS has finished releasing all their cross generation titles, this wave of titles since Starfield will all be designed around this console. Bars will be moved up, they’ve had a significant blowback from covid, sure, everyone did, but MS with the acquisition of ABK gives their studios more breathing room now that collectively revenue for Xbox is pretty high - they no longer need to build games designed just for monetization. MS can take significantly more risks with the types of games they put on their platform.
The fact that Sony landed barely a stones throw ahead of XSX in terms of compute and bandwidth sort of is the indicator that nothing else can be sold that is bigger/stronger at a reasonable price bracket for their customer base. That’s why we aren’t going to see a A midgen console for Xbox; why release a pro model with only 8 more CUs and a new architecture. I wouldn’t even know where MS would begin to try to market this.
Ideally if thry really had foresight of this, they should have sold chips that had a silicon profile that could handle the clocks and heat. And spend the last 4 years developing code to allow for variable clock rates and release that. Though the cost point I suspect would have been more extravagant thus very unlikely.
New hardware isn’t necessary, unless the intention is to shrink Series S further.
MS has finished releasing all their cross generation titles, this wave of titles since Starfield will all be designed around this console. Bars will be moved up, they’ve had a significant blowback from covid, sure, everyone did, but MS with the acquisition of ABK gives their studios more breathing room now that collectively revenue for Xbox is pretty high - they no longer need to build games designed just for monetization. MS can take significantly more risks with the types of games they put on their platform.
The fact that Sony landed barely a stones throw ahead of XSX in terms of compute and bandwidth sort of is the indicator that nothing else can be sold that is bigger/stronger at a reasonable price bracket for their customer base. That’s why we aren’t going to see a A midgen console for Xbox; why release a pro model with only 8 more CUs and a new architecture. I wouldn’t even know where MS would begin to try to market this.
Ideally if thry really had foresight of this, they should have sold chips that had a silicon profile that could handle the clocks and heat. And spend the last 4 years developing code to allow for variable clock rates and release that. Though the cost point I suspect would have been more extravagant thus very unlikely.
New hardware isn’t necessary, unless the intention is to shrink Series S further.
Johnny Awesome
Veteran
@iroboto I like most of what you say, but I still think MS could make a really nice box in 2025 for $599. The marketing is simple - "like the PS5 Pro, but twice as good." Sony is doing most of the marketing for them. I think it's a mistake to let the PS5 Pro go unchallenged for 4 years and a bigger mistake to launch their true next gen machine early, only to get outclassed a year later by PS6.
I think they should keep shadowing Sony a year later. That way Sony can suffer the software growing pains for a year with new hardware while MS never has to worry about lead platform issues because they generally have the more powerful box.
I think they should keep shadowing Sony a year later. That way Sony can suffer the software growing pains for a year with new hardware while MS never has to worry about lead platform issues because they generally have the more powerful box.
Last edited:
arandomguy
Veteran
I don't how that is realistic. It's already going to be tough selling the idea of the PS5 Pro as actually "being twice as a good" as the PS5 and that is with 4 years. Now you're asking Microsoft to come out with something for $599 just a year later and to sell it as "twice as good" as the PS5 Pro?
And I'm talking about this from an all encompassing perspective, as in whether it is technically twice as good, functionally twice as good, and also convincingly marketable as twice as good.
And I'm talking about this from an all encompassing perspective, as in whether it is technically twice as good, functionally twice as good, and also convincingly marketable as twice as good.
Similar threads
