Duderino

They have shown it off with Minecraft, a raytracing best case scenario. We haven’t seen evidence yet that MS + AMD have built a better raytracing solution. We know how their approach differs now, but are lacking adequate examples.

I don't doubt games this gen will get there, but unfortunately Minecraft is not indicative of the challenges other games will have with incorporating raytracing.

Can't say I'm surprised that Minecraft is their raytracing demo while Gears 5 relies on UE4's screen space global illumination instead of DXR. The cost of Raytracing BVH trees would be significantly less expensive in a game like Minecraft. In case you are not following, BVH trees are used to reduce the number of points each cast ray needs to test against. In the case of Minecraft, the world geometry is practically one big BVH tree already, simplifying computation and memory costs. In other words, keep your Raytracing expectations in check.

I very much doubt Sony would delay a game from a strategic launch date to ensure a simultaneous launch on PC. Answer me then, why are the Death Stranding and Horizon PC ports slated for summer and not early spring? With few AAA releases this quarter they could do even better releasing now. Doesn’t make sense, unless of course the work to bring these games to PC, going from a single console target, is responsible for the summer date.

AAA PC ports still require time and investment from engineering, QA, UX, design, etc. There is also no guarantee Sony 1st party game engines are even up and running on PC in the first place. Could be running straight off the devkits. First time they will be targeting nvidia cards too. It is a different situation from Xbox 1st party devs, many of which rely heavily on the years of cross platform development that has gone into UE4, unity, etc.

Maybe I’m in the minority, but I do appreciate how it differs from both Kingdom Hearts and FFXV. Especially for a Final Fantasy game, XV felt like a step back from the options, strategy, and abilities you get with turned based combat. I’m totally cool with the FFVII remake taking a step away from the A.D.D. combat (action air-dodge ) for something a little more measured and less out of DBZ like this: https://m.youtube.com/watch?v=cuH-6e2rK5U

There’s probably no right or wrong way to issue the attacks (like FFXV), but mashing one button will clearly result in Cloud taking more damage than necessary. I get the impression that the general melee is lite in the way of required button dexterity for a reason. Managing a third party member Is likely going to be demanding enough.

I've heard far worse criticisms of the original FF7 back in the day. The player drop off rate after an hour or less of play was probably astronomical, at least in the US.

Sure, if your friend was pocketing $5 a month to loan his hardware to you. Comercial software or services that allow you play (or view) content on another potentially unsupported device can be pretty good grounds for a lawsuit if there is no agreement in place. Even if nVidia were to win the potential case(s) in court, the service could be taken offline until the verdict is reached. See what happened to Connectix. ( How ironic it would be to see Apple sue nVidia over this )

After playing the demo I can't help but think the complaints about the combat in this thread are a bit on the hyperbolic side. First, it's not a button masher. Yes, you can needlessly bash the attack button rather than holding it down to commit attacks, but when battles pick up with the more dynamic enemies excessive mashing would be a poor choice. Second, despite some surface level similarities, this is a good departure from FFXV's combat. XV required nowhere near this level of player input, respose, party management, and positioning. My first impression is that they made a pretty engaging combat system. Just might not be everyone's cup of tea, at least at first. Looking forward to seeing how it holds up with a 3rd party member.

So if I was ok with FFXV’s combat will I be ok with this? Sounds like the same deal where you hold attack and block on occasion, then trigger special attacks when ready. Would it be wrong to say FVII:R’s combat is a bit more varied and tactical? (A Good thing)

Looking like it’s quite plausible for a modern 4TF-ish GPU to keep pace with the XBoneX’s 6TF. Maybe even double the FPS if the goal of a potential Lockhart is 1080p-1440p. Question remains though, would a GPU close the XBoneX’s output be enough to keep pace with the next 6+ years of advancements in shaders/lighting/FX, GPU assisted simulation + other compute tasks, and rising assets counts and costs. I suppose that’s why the developers Jason Schreier spoke to had reservations.

Digital Foundry have speculated the same. Not expecting either console to have dedicated, Nvidia style, tensor cores for BVH and denoising operations, but I do think in other ways the hardware will be tuned for it. At least to a lesser degree. More speculation here, but I suspect Raytracing on consoles will be paired back in the way of: Less cast rays Lower quality denoising Lots more temporal smoothing Exclusion of many or all non-static objects Reduced raytracing distances, missing objects in reflections, lower raytraced LODs Less ornate geometry that would be expensive for rayracing to accurately represent. Less use of "full" raytracing (shadows, ao, global illumination, and reflections). Continued reliance on screen space techniques, cube maps, voxel based methods, and/or light probes/baking to make up the difference.

Going by varying performance of DXR enabled tittles on PC there is little reason to believe Microsoft has universally solved the raytracing optimization problem. The best examples today are just as much a testiment to each dev team's own advancements. In all likelihood that will also be the reality with next gen consoles (with Vulkan too). Sure won't stop marketing from doubling down on Raytracing though, even if at the start of the gen it's not common place.

Calling it now, there is going to be quite the performance/quality gap between developers that build custom RT tech and those that do little more than rely on what comes with DXR. Furthermore we’re going to see frame rates take a dump in a lot of unexpected ways.

In light of the recent rumor that Cyberpunk 2077 was delayed due poor performance on the PS4 and Xbox One, I do wonder if we'll see a few MS first party studios struggle to keep their ambitions in line with the base Xbox One hardware.

Getting flashbacks of Bungie porting Halo: Combat Evolved to PC a year before Halo 2’s release on console.

I wonder if Spiderman crashing into a spawning cars, billboards, and water towers would sit well with their audience.

With disc HDDs, one of the optimization tasks developers have to deal with is grouping data together into packages. In other words, store assets that are frequently accessed together close on disc. For example, there might be an environment biome package, an NPC faction package, a regional audio package, etc. The goal is to reduce the distance that needle needs to travel to access the requested block of data. When assets do not breakdown neatly into these separate groups, there are 3 options: 1. Pre-load assets into global memory if space is available. 2. Duplicate assets across multiple packages on disc. 3. Suffer the potential performance consequences. (Option 2 is more common than you would think.) The HDD disc is a limitation that can impact everything from the overall game size to how worlds are designed. A developer could of course build a game around the PS4/Xbone HDDs, but their next-gen ambitions would be restricted by it.

Something else to consider, 3rd party cross-gen games will need to support PS4, PS4 Pro, Xbox One, Xbox One X, PS5, Xbox Series X, likely PC, possibly Lockhart, and maybe even Stadia. That’s an extremely tall order that I very much doubt will be sustainable for 2-3 years for the vast majority of developers.

The context of what the GPU has to render factors into those resolution costs. As that complexity increases next gen, so will the challenge of supporting less capable GPUs whether it’s the last gen consoles, older PCs, or perhaps now Lockhart. Not everything that negatively impacts GPU performance is guaranteed to have a convenient dial, a low acceptable threshold, an available alternative method, or a clear and easy path forward with optimization. At least on PC devs can bump up the minimum and recommended specs.

Both Jason Shreier and DF are clearly talking about Lockhart in the context of the Scarlett family. Not the Xbox One. I have doubts too, but those have more to do with the practicality of Lockhart than the validity of these rumors and their sources.

I trust that Jason Schreier (Kotaku) has accurate sources, that Digital Foundry has reason to believe it is still in the works, and well... that an outspoken Jonathan Blow (the Witness) is not looking forward to at least one new console that is not the PS5. Lockhart is by no means a forgone conclusion, but there is enough here to have a discussion about it as a possibility. In the case of the X, it's a lot easier to scale games up. Scaling down next-gen tittles for a hypothetical Lockhart would be a bigger ask of developers. Much easier to work with less limitations than more. I doubt Microsoft wants to add more work for developers and possibly introduce limitations, but a second SKU is not outside of the realm of possibility if it fits their consumer facing strategy. Putting all problematic PC ports aside, I would agree that architecture is not the potential challenge with Lockhart. The power differential is. That's one optimization strategy that could help when resolution is contributing to the bottleneck. Different story though if the key limiting factors come from draw calls, tri counts, and/or compute shader performance.

UE4 has been a mixed bag this generation in terms of scalability, at least as it relates to performance. On one hand Fortnite, Gears 5, MK11, and others seem to scale well. On the other, you have Jedi: Fallen Order, PUBG, State of Decay 2, We Happy Few, and many more that struggle on the base Xbox and/or PS4. It's not just Unreal either, look at Anthem, Control, etc. At least the Lockheart rumored CPU should be less of an issue but we're still looking at a GPU that is closer to the Pro or X. Even with a target resolution of 1440p, higher poly counts, increased shading complexity, and more demanding games in general are likely going to result in Lockheart requiring special considerations and optimizations. Possibly even separate lighting solutions from the Series X or PS5.

Or vise-versa. If the PS5 is the worldwide market leader and becomes the primary dev platform for many, Lockheart ports could very well receive the least care and consideration. The console would have to sell exceptionally well for it to make more of an impact.