I don’t think arguing that SVOGI, or any RT GI in particular, is something redundant, has any base under it at all. It is warmly welcome feature.
Is it that absolutely needed? I don’t think so.
For example, I would definitely favor seeing landscape rendering in UE4 fully overhauled instead of seeing dynamic GI in UE4 in next year or two.
When you talk about performance numbers don’t use FPS: It does not tell a thing. Difference from 30fps to 27fps is 3.7ms. From 60fps to 57fps it’s 0.877ms. Always measure how much in absolute time some feature use. It’s not easy to calculate (1000/oldFPS) - (1000/newFPS) in your head and if you only tell the difference you can’t even calculate it.
VXGI is the direction that Unreal is heading, which is fine because I’m assuming they worked a deal out with Nvidia. Both SVOGI and VXGI use similar methods, the latter just uses 3d clipmaps instead of octrees. The pros/cons between the two methods vary. 3d clipmaps(VXGI) will give you better performance, memory usage, and handle dynamic objects faster, but you will end up with varying density aka varying quality(stuff in the distance won’t be as accurate). They just need to finish working out some of the kinks to it involving tracing/blending to prevent flickering due to the changing densities in the clipmap. I’m assuming that’s where a lot of the performance hit resides at the moment.
Here’s an extremely good writeup comparing the two: https://erkaman.github.io/img/master…thesis_101.pdf
If you look toward the end, you can see some performance comparisons.
And I’d definitely agree with you on the landscape part…
No. If some they want someone for the Fortnite, they add this info to the title of job offer.
Check this for yourself, I count less than 10 offers specifically for this game.
https://epicgames.avature.net/career…s/?jobOffset=0
Of course, some of the engine guys spend time on engine features required by Fortnite. Still, they merge a lot of good stuff from their projects into the vanilla engine.
Question is: how many engine systems wouldn’t exist at all if Epic wouldn’t create their own games? I doubt that we would have skeleton animation tools and AI if they wouldn’t develop big games that need these systems.
Seriously they don’t need all these engineers to create skins for Fortnite 
Yes, I was looking at that page; this is why I edit that post.
However the majority of jobs there ARE for Fortnite related things (backend, account security, cheat exploits, DDoS protection, community managers, QA, translators, etc, etc)
I didn’t see any “Graphics Engineer” position btw.
What about Rendering Programmer? 
https://epicgames.avature.net/careers/SearchJobs/?jobOffset=90
Yeah, you’re obviously right about community guys.
Although they also recruit people to work on engine’s community. We would definitely use better communication with Epic. Hiring lot of people taking proper of care forums, requests, questions - it would be so good! The community grew way too big to be handled by engineers a long time ago…
https://epicgames.avature.net/careers/JobDetail/Cary-NC-United-States-Community-Manager-Unreal-Engine/2714
Oh yeah, overlooked that!
Let’s hope they find the people they need and something good comes out of that :s
Natalya Tatarchuk, did you see how much she’s done already for Unity in such a SHORT period of time??
My god, Epic need someone her level.
suddenly there’s no manpower budget for rendering features, and yet we keep getting updates to the VR editor, VR/AR/MR, an entire ‘suite’ for Archviz, etc?
issue is UE4 tries to cater for a too broad audience now, and the classic core games audience (what used to be the audience in UE3) is not as significant anymore
Maybe that has to do with the fact most people playing that kind of “core” games… they play games where Unreal isn’t used anymore ?!
I mean, all those AAA games now have their own AAA game engines since mid 2000s; many barely license UE4 to use it’s source code as a skeleton and go from there (never buying a Unreal license ever again). So engine manufacture must go where potential moneiz are, I guess.
CryTech is focusing on their “niche” engine users because most their latest released games kinda bombed, they NEED CryEngine to survive so I see them focusing a lot on engine features lately.
Unity with this 2018 revamp are starting to do the same as CryTech, are giving a break with that “cloud services” BS, focusing on “game development” finally (this is weird rofl, a game engine is finally trying to act like one).
I know that Disney injected money in Epic Games, this is why they came up with that “RTX” thing, Disney need that on their VR attractions in their theme parks.
A few major games such as Final Fantasy 7R are using UE4 as well, they are totally fine without GI or open world features.
Tesla and few other car manufactures are using Unreal 4 to design cars.
An ocean of VR startups are using UE4, I did some freelance work for a few btw.
Then what’s left?! Steam indie devs?
I think this is the only landscape Unity or CryTech are competitors to Epic; might be reason why they seem not too much invested to this market, everyone’s noticing how features these devs need aren’t priority anymore 
They have. It was never problem of people. It is a problem of of where resources are allocated.
Epic just announced today a tournament prize pool of U$100 million to establish Fortnite in the eSport scene…
No eSport to date ever had a prize that large.
Really, engine is far from being priority anymore. Expect this company to spin a 900° changing a lot upcoming years… There’s already “League of Legends stars” migrating to Fortnite and videos around saying Fortnite is killing LoL.
After eSports houses such as TSM know about that prize Pool they won’t think twice changing focus over to Fortnite; there’s soooo much happening around Fortnite that someone very powerful involved with this might tell Epic to stop changing the Game Engine for the sake of optimizing Fortnite for the upcoming tournaments.
That could result on a stagnation of UE4 soon, effectively turning it into the “Fortnite Engine”.
I guess we should move our discussions and bug reports to fortnite forums to get Epic more involved 
Time to Jump ship, question is where do we all go?
That’s not how it works for the ppl I’m working for…
4.19 is very bad version, but still usable;
If from 4.20 the game breaking bugs aren’t fixed then we will move back to 4.18 and after the game is complete I don’t know what they’ll decide to do.
The top cards in that list are at same performance range of previous generation.
Asking PC gamers for a GTX 1050 or better for recommended specs isn’t unreasonable for graphically intensive game.
Remember there’s a ton of Steam users that play mostly DotA, CS:GO, etc, not demanding games that deflate the average. And those gamesg aren’t your target audience as a developer because they stick to one core game.
Crysis 3 is all about realtime lighting / GI… There are quite a few videos on youtube showing people playing it on i3 + 1050Ti. So, the so called previous gen GPUs work quite well with high fidelity AAA games that have real-time GI. Obviously engine has to be optimized well.
I played through Doom 2016 using 670GTX and AMD Phenom x3 (ancient CPU). While it doesn’t have real-time GI, it has a ton of shadow casting realtime lights. I doubt I’d be able to play the game if it was UE4-built game.
Yes unfortunately UE4 is very different.
That’s what I was trying to say… Not saying it’s bad optimized like older versions of Unity, but UE4 is not CryEngine. It wasn’t built around real-time GI and even CryTek didn’t get fully get there yet after so many years.
You add GI to a serious UE4 game and those GPUs will probably melt down.
I don’t mean to disappoint you, but there is nothing in UE4, that would make its dynamic shadowcasting slower, than elsewhere, including IDT6. There are few cases where things could be traded off here and there yeah, but when speaking about local shadowcasting lights, the tech is the same. It is down to per-project tuning.