Discussion RPG - UE5 vs Unity - RPG Open World

Hey there,

i know there are many videos out there about this topic but none of those really help out.

We have been developing for several years now in Unity and it was a very nice journey but it seems like for "Large scale" Games, Unity is having bottlenecks as well as there are not so many (InEngine)Tools there.

I have done a Pros and contra list (out of my experience, not everyones ofc.) where i have written down certain Points which seem to me important for Games.

PLEASE BE AWARE, thats the experience i have made with Unity ofc. i dont want to say that Unity is bad because !its not! but maybe it not the best choice for Open World RPGs with a large amount of crowd but for many different games?.

Unity Pros:
- Great Community
- Beginner Friendly (Doenst apply to my case but its worth to mention)
- Great Docs
- Offers a small range of templates (Again in here, doesnt really help out in a bigger commerical game, worth to mention)
- DOTS (Starts to rise)

Unity Cons:
- Monobehaviours: ("very" Expensive but also here, the way Unity handle Monos is good for beginners but its also consuming a large amount of resources, ofc you can avoid using Monos but that here is not the point)
- Terrain Tools: (Sadly we have to buy Assets to improve the Terrain possibilities as for example, making Caves, Water oceans or rivers etc.[Or ofc. code our own API]) i know that they are working on an Water API for example but its still in development.
- AI Navigation: (Its quite out of date sadly, no updates coming since a long long time, here we have to go back to A* Pathing which offers a really great support, docs and a large amount of functions even for Open World)
- AI Tools: (As for example a State Machine, i mean its not too complicated to create a FSM or a Behaviour Tree etc. BUT its still desireable to have this in the Engine as for example UE have)
- World Streaming: (Well here from Unitys point of view we have to implement our own logic, which is "time consuming" but doable ofc. still would expect something here from Unity in terms of Open World games and overall it helps also in small games)
- UI: (UI Builder is coming but havent seen an Update for over a year here [please correct me if im wrong but looking at their git repo, nothing changed), besides UI Builder, the UI Toolkit is out of date and also here, there are many nice Assets which solve that issue but again here, you have to pay for something you would wish to have in your engine instead of having to pay for that and also not to depend on other assets then you do already)

Unity Unknown:
- VFX: I know we have the Shader Graph here, but Unity doesnt offer any Effects out of the Box which are usefull, as for example fire or water etc. (To my knowledge ofc!).
We can do alot with the Shader Graph but also here, there are certain things which i would wish to have them out of the Box with minor adjustments

Again, Unity is great we have been doing very well for the past years but we are considering to switch to UE 5 for the sake of an RPG.

UE Pros:
- Since UE 5 the Docs are great
- Great Render out of the Box
- C++ (could also be a disadvantatge to whom are not mighty of C++)
- Offers HUGE Templates with MANY features and great best practises
- Great Terrain Tools
- World Streaming is amazing
- Amazing Lighting
- Auto LOD (For Components as well es for 3D Elements, i think everyone saw the Demo )
- Offers AI Tools as for Example State Machines
- VFX Offers Water, Fire and many more with stunning graphics

UE Cons:
- Smaller community
- C++ (Which can be a Pro/Con as mentioned above)
- GUI (Can be hard to follow for beginners)

UE Unknown:
- AI Navigation (Didnt went to deep into that but it looks like a very detailed, performant with alot of possibilties) Navigation
- UI


Would also mention that we speak here, about the Engines in case of an RPG.

Also small Quote:

Industry Presence
You may choose a game engine based on what the professionals are using. Both Unity and Unreal are used to create games on the market, but in different ways.

First, Unity is the most popular engine for indie developers and mobile games. There are a number of larger games made with Unity such as: Hearthstone, Cities: Skylines, Rust, Ori and the Blind Forest, and most mobile games.

In terms of the AAA-industry, Unreal is used more than Unity. Games such as: Fortnite, Bioshock, Sea of Thieves, Star Wars: Jedi Fallen Order, and a large number of others use the engine.

Something to also keep in mind is how the engine developers themselves use it. Unity don’t create their own games apart from small educational projects. Epic Games (developers of the Unreal Engine) on the other hand, have developed many games such as: Fortnite and Gears of War using the Unreal Engine.

AT the end, Unity is great, also Unreal is great but in terms of great graphics and Open World Games with a large amount of crowd, i see many difficulties in regards of Unity (DOTS Help 100%) but its no out there yet for PROD althought there are "Amazing" Games done with it


I Hope everyone could follow my Topic, and the question woudl be:

What are your experiences with UE compared to Unity in terms of Open World Games?

 

I'm not sure this belongs to the scripting forum as this has nothing to do with scripting. I think this belongs more in the general discussion sub forum.

Anyways, a huge difference between Unity and Unreal is that Unreal like most other engines have been developed for a certain game / game type. So most of the tools they have implemented were required by their own team to make a specific game. Of course as unreal has followed Unity they also added some more general things, but in the end it always was a quite focused engine.

Unity on the other hand is designed to be as generic as possible and is highly modular. The unreal argument "Great Render out of the Box" is purely based on content which they already had from the games they made and decided to share / include with the engine. Unity has the assetstore which provides the equivalent to build-in resources of unreal. Of course, some are paid assets some are free and it may not be the easiest task to find the right things immediately.

Though Unity is much more lightweight and customizable with its modular approach. I don't see a reason why for example a huge AI or navigation package should be included in the engine. Most games made with Unity are 2d mobile games. There you barely need much AI or terrain. If you need you grab the tools you want or create your own. Easy tooling makes Unity quite strong.

I personally have not really used unreal. I played around with UnrealEd back in the days of the first Unreal Tournament, but that doesn't really count

"Open World" is a really vague term. There has been MMOs out there made with Unity like Battlestar Galactica Online. Though sadly the servers have been shut down after 8 years (3 years ago). Here's some footage. Actually a non-profit fan based community are currently developing their own servers (with permission) and partly revive the game.

I think in Unity you have much more freedom of choice. Yes, some features would be great if there would be an out-of-the-box solution, however Unity is pretty much genre agnostic.

The argument that Epic made a lot games with their own engine doesn't really mean anything. Of course they did and as I said, that's the main reason why there's more build-in content. Developing a game is never easy. So having an engine providing a lot pre-made stuff is not really a good argument. You should focus more on teh actual features. Again, it highly depends on the type of game you want to make and what concrete features you need for that.

 

Really great comment, thanks alot for sharing your experience and Battlestar Galactica, will take a look at it, looking for more of such Informative comments

 

Play around with both engines and make a choice.

 

Time is a certain faktor here, thats why i hope for some developer experience.

Started already with a few things but not ready to judge

 

You are starting an open world RPG, which will probably be a multi-year endeavour and you can't spend a month to properly to evaluate both engines and would rather let randos on forums make the decision for you?

 

Thats not right, i am searching for opinions and experience of people who might tell me what they went through and if UE is a benefit onto my project, im not letting anyone decide.".

Its better to listen to someone who went through this process is ready to share his point of view. That will increase my productivity in that regard as well

Dont missunderstand the question pls ..

 

Opinion: Don't use Unity.

 

Couple of random thoughts about these things and why I find this whole UE is the MF man on the block and Unity is supposed to be hot garbage very misguided and mostly false (there are some valid points, like asset streaming)...

Why is it not the point? If you don't like the overhead, you have the option to limit MBs to the absolute necessary, you even have custom player loops built-in to help you with it.

I'm sorry, but have you seen the thing UE calls "UI editing"? I wouldn't put Unity's existing and in-house solutions in the PRO category, but wouldn't put them in the CON either. You can make it work with relatively okay time investment.

Yeah, it's called Visual Effect Graph and available in the Package Manager. There are also samples you can utilize, but making new effects through the editor is fairly simple, and there are countless sources where people can give you ideas how to achieve things. For Built-in rendering here are sample assets.

You shouldn't, unless you're matching the number of people working on your project and the amount of budget you have. You should choose your tools for yourself, what you can achieve with them. Your team isn't any of those teams.

Also from GTFO to Hardspace: Shipbreaker, from Kerbal SP to Escape from Tarkov.

Unless you're an AAA team (with a lot of people, high budget, no desire to invent new ways of gaming and storytelling) there is no relevance. If you're, Unity isn't for you, probably and Unreal is a middle-ground.

This is a big myth (BS). The engine developers at Epic use exactly the same amount of Unreal on average as Unity engineers. Engine developers are developing the engine, not a game. There is a caveat to this though (see later). But, the people who are working on games are talking to the people who are developing the engine. The same is true to Unity. Not developing a game in-house means nothing. Unity engineers are on game projects at clients by the dozens. They are called success-engineers or whatever and their job is to visit clients and solve hard problems for them (and report back to the engine team about their findings). Since most of Epic's game development is done by outsourcing I don't see how it would be any different.

Again, no. Not the same people, majority of them aren't even working for the same company.
And the caveat is: it is true that the engine developers are listening more to the game developer teams at Epic on the short run, but this has an unintended side effect: if you make games what Epic already made, you will have an okay time with the engine, maybe even good time. But when you try to deviate from the path of the generic AAA-style FPS/TPS games, you will have a much harder time. Also if you want to invent any original game play or whatnot, Unreal's advantage is at least partly gone. Epic is making Unreal for their own projects and the wider audience is secondary. Unity is making an engine to primarily everyone (this has side effects as well, just check how many platforms Unity is available on).

So yeah, if you intend to puzzle together another Fortnite, Bioshock or GoW game, Unreal is the way to go 100% without a doubt. If you want to make something different, you should think about it what your team can achieve with the engines you have knowledge of and which is the more appropriate for the job at hand.

 

Nice one. Also might add that from my experience performance optimization is a bit harder with unreal, but might also have used the wrong tools for unreal since I've only worked on 2 small unreal projects for performance

 

One additional factor you could consider is "politics". There was a phase of more than a year where there was zero communications about DOTS-ECS. No new versions to test, absolutely no communication on the forums. So the "early adopters" were clueless wether it's wise to invest more time or the the whole thing is scrapped already. And since none of the many staff talked any more it's safe to assume this silence was ordered. Should a company treat their "enthusiastic" users this way?

Another point is, that UT is acquiring and/or merging with other companies like a street "lady". They do not focus on "low level" game dev any more (at least it seems to me) but they want to play in the shark tank of high risk capitalizm. So the only thing that counts is to bloat up the stock value. And since providing a good product doesnt help the stock value at all, paying customers have low priority.
Just consider what could have been done with the billions used to for those "acquisitions". They could just license some of those "packages" you need (terrain, AI, Graphics) and make them available for free (like they did with TextMeshPro). No. You have to pay (and UT is receiving their 30% share) for this stuff which COULD be there already.

And this whole DOTS/ECS is another chapter. There is no 2d solution planned. Animation, Sound are virtually not existent from what I read (some users provide their own frameworks). This whole thing is more than five years in the works and still in experimental. Keeping the GameObject stuff for "authoring" (called hybrid) was a wrong decision from my POV. I tried to use ECS some months ago. Had a showstopper bug in the engine. Once it was fixed I ran straight into the next. There was no help in the forums from Unity staff. Probably the bug is still around. I could not advise people to use DOTS-ECS right now. And I'm not sure this will change in one or two years.

For me the "management" has deviated from their core business (and promise) to "democratize game development". So you should not just compare feature sets now, but also what you think both engines are heading towards. If you choose Unity you can bet that they have acquired 10 or 20 more corporations or "services". If you choose Unreal you can expect they add more usefull stuff.

I have never used Unreal so I can't really give advise which to choose feature wise and for your requirements. But lately I regret having choosed Unity back then. They were the "congenial underdogs" from europe providing an indie friendly engine. Now they are capitalist overlords ruthlessly fighting for market share and stock value.

And I'm not a fan of Epic either. They have also done their fair amount of stupid stuff. But the money you spend supports those "politics" and management decisions. If you don't like them, vote with your wallet.

 

I will point out that a lot of the people that started using Unreal and found it to be better than Unity are probably not here any more.

So by asking here it seems you have made your decision (unless you started a similar thread on the Unreal forums?) and you just want some validation? In which case, read any message other than mine.

But again, I urge you, take a month and evaluate both engines as much as you can, if you are starting a longer term project, surely one month is not a significant amount of time.

Good luck.

 

Thank you for your time, althought i disagree with 1-2 points since i workat InnoGames and we do games with unity since i get here, so about 5 years already, i do understand what you mean.

 

Also here, thank you for your point of view.

It is actualy true, to look not only at the tools provided but also "where the company is going", thanks for that one

 

it seems to me that you speculate and judge without me pointing any of those points out.

Just let it be, no need for that

 

It's not speculation. If you've been with this community for as long as we have you start to memorize the names that you run into on a regular basis. Many of the ones who are no longer here are now only in the UE community with very few exceptions. @BIGTIMEMASTER is one of the exceptions.

 

By speculation i dont mean that point but the one saying that "i have made my decision already" which is not true, as said, im hoping for some good information as stated above from the previous people who took their time to write down their thoughts.

Its like reading a book and hoping for good informations

 

Speaking of which I recommend creating a thread in the UE community if you haven't already.

 

Havent done that yet, but a very good idea. Thank you for this one, will do so as well

 

I think you should update thread topic, to highlight, that you address only large world with floating point challangea and RPG type of games.

As already pointed out by others, your discussion doesn't apply to all game making scenarios.

This basically invalidates the OP discussion, as if Unreal is better than Unity, as it is suggested in OP. Unless focusing on narrow use case.

But saying that

Unity was for long time handle large amount of dynamic instances like crowds. City Skylines is a good example of many.
Now with DOTS, even more possibilities are open, and people do release games, with high count of dynamic instances.

So for that reason, I don't agree with your point on big "crowds" .

 

I spend time between each project evaluating the available tools. I recently spent time comparing Unity 2021 with UE5 for the purpose of making an open world game. Both engines have their advantages. I was specifically comparing Unity 2021.3 with HDRP vs UE5 with Nanite and Lumen.

There are some good assets out there for Unity that handle things like world streaming. Gaia Pro 2021 even includes world streaming options, so you don't need an additional asset to manage that. If you want a separate streaming, another option is World Streamer 2. There are good weather and time of day assets available for both engines. There are also grass rendering options available. For Unity, you will need to get something like Vegetation Studio Pro to handle GPU instancing of the grass, plants, and trees. Rendering vegetation as game objects obviously won't scale.

I personally chose Unity 2021.3 with HDRP for my current project. One big part of my decision was how productive I am in Unity vs UE5. I have a lot of experience with Unity, so I am personally more productive with Unity. Additionally, UE5 looked great with Nanite and Lumens, but my workstation was often very bogged down by UE5. Everything took a lot longer in the UE5 editor. My workstation is an AMD Ryzen 3900X, 32GB RAM, Nvidia RTX 2080ti, and multiple NVMe drives. That is plenty of computer to operate Unity, but that is basically minimum spec for developing with UE5/Nanite/Lumen.

 

Have a nice life.

 

There aren't a lot of RPGs done in Unreal Engine. Likely for a reason.

Basically, RPG is highly likely to feature a complex UI, and I've never seen any unreal game display anything remotely similar to Pillars of Eternity/Tyranny journals/log screens where you could hover over individual words and see hints. With unity UI, which you listed as a "con", you'd have no problem making pretty much anything, even though the process is definitely going to be awkward in some places.

You should be able to make an ARPG though.

Regarding pros and cons...
* Unity offers very clean API (which is a pro), and Unreal offers very messy API (which is a con) and forces you to use blueprints. C++ is nice in theory, in practice AFAIK Unreal uses a dialect(UPROPERTY says hi) with exceptions disabled, and the engine is not easy to extend.
* Monobehaviours are a pro, as they're far easier to prototype with.
* Unreal does not hav Mecanim.
* "Open world Games" does not mean "RPG".

My opinion is that if you get distracted by unreal world building tools, you'll bite off more than you can chew by t rying to duplicate an AAA, then your project will die, because that'll be above your abilities to produce content.

 

Game project will take several years to complete.

Invest a month or two to prototype in each engine. Stress test for project bottlenecks.

most people have not made an rpg. Many have not even made a commercial game. Opinions and extrapolations count for tee-tee. Only way to know is to test.

 

Okay, if you work with Unity already, why are you asking us? I'm just intrigued by this question you seem to ask. You say time is of the essence, it probably isn't unless you have some publisher saying, "make this game in 12 months." Just make the game and stop asking questions. I've seen people here for a decade and not make a single game. Meanwhile I have made and released almost a dozen! Go and create!

 

I definitely want to try Unreal Engine one of these days. None of the "pros" that you listed for Unreal really apply to my project, though. Still, I want to experience what Unreal has to offer vs. Unity. I chose Unity for one reason and for only that reason- that is their free version came first.

I'm working on an open-world game too, and (in retrospect) I'm glad I didn't really worry to much about which engine facilitates streaming content or better rendering, because really I would have just been wasting my time. Turns out my big big "bottleneck" is actually creating the content, and that has nothing to do with either engine. I could keep modelling, prefab-placing, writing, and puzzle-designing for the rest of my life and still never reach that too-far-away-from-the-origin floating point limit.

Neither can I efficiently create high-quality graphics assets. I know how to make metalness/smoothness maps and normal maps and such, but it takes too long when you add everything up. I ended-up choosing a simple, retro style for my graphics that would easily work in either engine.

I guess what I'm saying is - are you sure that you really even know what you need at this point? I did not when I started my project. Maybe you are working with a big team of industry pros (that I haven't heard about) and they already have an idea about what their content pipeline is going to look like.

 

Thanks for the advice in terms of title change, i did change it, thank you

In terms of "Big cowd", yes it is true that with DOTS it is more than possible but its lacking of features for me, for now as in the answer from @exiguous described.

I Agree that i could write my own (for example) Animation Controller for Animation transitions and so on, but everything takes quite alot of time already besides the fact of creating an RPG. I know i should be willing to spend time for that as well but i am still searching for the best tool to work with.

 

Very much appreciated, that answer, put alot of informations down.

I have all assets mentioned here aside of World Streaming from Gaia/World Streamer. Will take a look at it.

We have, as mentioned above alot of experience in unity and in terms of creating a game we have already our "Tool Box" which includes many APIs to create a game quite quick. And here comes the point, in terms of a large scale game with alot of crowd we will have to put hands on into DOTS for sure, is it worth more to learn DOTS or UE(from basic experience). I would say Unity is the go here.

I will get the World Streaming Package from Gaia and see how it goes.

(Thats the only point which makes me feel not so good about Unity, we have to access for a hand full of "very important things in my point of view" the assetstore and purchase those) Also here it can happen that you buy something and it is not properly optimized/done which could cause issues in the future)

Will defenetley come back here with a feedback, thanks alot for yout time!

 

Hey, thanks for your reply!

I Agree with most of the points but the RPG one.

In UE i see way more RPGs gettin done than in Unity at all. Maybe its my fault that i did not clarify what game genre i mean by "RPG". I mean games like "Gothic 1,2,3, Risen 1,2, The Witcher, Skyrim, the new Harry Potter RPG etc". (I know that not all of those games are done in UE but its just to make clear what genre in general i mean).

Please correct me if im wrong, maybe i did not a good research on that.

 

Great Experience here.

I did an RPG Prototype with High Quality Graphics for around 12 Months. It looks amazing, its great, but i start seeing "needed" things like a more dynamic World Streaming then i have prototyped, terrain manipulation as mentioned above "caves" here for i have done a basic tool to make it possible, there are assets in the store which does that for me but its again another external dependency which i dont know how it behaves in certain circumstances. Dense forest does it alot of performance including large crowd around. Speaking about large crowd, i have running over 400 units at once with a high quality world but i see them eating quite a good amount of performance here for i would have to work with DOTS for sure(which is a bit of a problem since i am missing tools here as well which i would have to create again, im sure they will come since its still not out for PROD but its still a faktor).

But its really nice to hear that you made such a good experience, will try some things out and see how it goes.

 

While not all the same style (some 2d, pixel, 3d like the ones you said, etc), 2300 made in unity.
Top Role Playing games made with Unity - itch.io

180 in Unreal.
Top Role Playing games made with Unreal Engine - itch.io

Some games made with Unity.
Made With Unity | Unity

 

Looks like a large list, but i would not count games which made under 100k revenue to my list.
Dont understand me wrong but there are "WAY" to many games made with Unity which are (i dont wanna be rude) not good at all. Already on Steam there are (i think since the last time i checked) 7x more games made with Unity BUT 90% of them are not good (at least in my point of view) and ofc. please dont think i mean to be rude, i dont, i just want to share my point of view

Again, thats on me to clarify my point of view which i havent done above.
But still, i see at the link "made with Unity", those look really great.

 

Well there are top level games that are made with Unity such as Subnautica, Genshin Impact and Pathfinder Kingmaker (looks like they didn't make into these lists).

My opinion is that the strong point of UE5 is their megascan library, if you want to add realistic rocks and trees to your game then that's the way to go. If you want your game to be moddable then UE5 also provides better tools for that.

However it's not the engine alone that makes great games, it's the game designers.

For me it was more about picking C# vs C++ and I was more comfortable with C#.

Jason Weimann has a couple videos on YT about Unity vs Unreal if you want to watch them.

 

Plenty of unity games have made that money.

I don't mean any disrespect though, but if you're only looking at the money aspect here, you're thinking very wishful. You need to make a game first to make any money and your first game probably won't make much realistically.

I think you're spending a lot of time asking questions when you could've already been trying both engines the past few days.

 

I’ve only skimmed over this thread so I might be duplicating what's already been said. I use Unity at work. I’ve spent some of my spare time on Unreal but I’ve never released a game with it.

Things I like about unreal.

• Better graphics, more artist friendly
• Superior multiplayer support
• Free assets (not that I’d probably want to use them e.g. sonic frontiers). Infinity Blade $3 million, Paragon $5 million

Things I dislike about unreal

• I prefer c# over c++ and blueprints.
• I found it takes a lot longer to build content in Unreal.
• The cost of realistic graphics skyrockets so I normally prefer more stylized graphics (toon / pixar).

Another factor is the licensing e.g. Unity per seat vs Unreal 5%. As @dlorre mentioned, Genshin Impact is a great example of an open world ARPG Unity game. If I had a budget greater than $10 million I’d probably use Unreal, otherwise I’d stick with Unity. I'd say that Unreal games tend to have a higher quality due to the larger budgets.

 

Wait for Witcher 4 on UE5, they will then have a toolset for RPGs (if I am not wrong they, CD Projekt, agreed to share improvements to the engine...correct me if I am wrong)

 

Great point, its not about the money in first place at least and i have been working already in a succesfull company here in germany, working with Unity for several years BUT we do only mobile games which are very successfull, therefor already did an prototype in Unity which took good 12 months with everything quite good working but still certain points are bottlenecks i would say. Im touching a bit of UE atm but cant give a real summary since i need to dig into the engine alot more.

 

Thats one of the Points i really like about UE (at least from reading if it is true), they do take many systems to UE over from different Projects and open them up to their community for use

 

I heard that one here alot:

• I found it takes a lot longer to build content in Unreal.

Which is a big faktor (to me) and to many others i would say.

Thank you for sharing your Experience

 

The Gaia asset with world streaming support is "Gaia Pro 2021". It can break a large map into lots of small sections, and then load/unload those as needed at runtime. It also includes an option for automatically working around the limitations of floating point. All of the popular vegetation rendering solutions also have a similar option for dealing with floating point, so you can sync up that option if you need to deal with map sizes that would cause issues with floating point.

As far as DOTS goes, I love that tech but I would not plan any projects around DOTS yet. There are ways DOTS can be used for various parts of a project, but DOTS is not ready yet to handle everything in a project.

 

Yes, i have done a small World Streaming solution, just for prototype purpose, its crazy how the performance improves by that, i really like that.

Regarding DOTS, well i would not use the "whole" package but using Burst for expensive Actions will be a benefit for sure.

We have atm with the MonoBehaviours around 400 Entities Runtime walking around and its still a chunck of consumption. What i want to do is, disable everything but the Transform which makes the AI move at least, so if the player at a certain daytime comes close, the AI will be there where it should be.

Aside of that, there will be more AI for sure and i have heard in one of the last Unity Videos that a Studio delt with a simular problem. One of Unitys Teams helpen out using the Burst compiler to do what i have described and it improved their performance by far. So i will have to see how far i can deal with the Burst compiler at least to work clean.

Thanks again for the top info!

 

By RPG I mean CRPG. Planescape: Torment, Baldur's Gate, Neverwinter Nights, Pillars of Eternity, Tyranny, Shadowrun: Dragonfall.

In Unreal there's pretty much nothing, as I can't remember even ONE such title made in Unreal engine.
In Unity engine we have Pillars, Tyranny, Shadowrun, Wasteland 2 and Wasteland 3. It is proven and battle-tested for this genre.

I do believe t here's a problem with unreal gui, as it is very uncommon to see anything complex done in it, while in unity UI you absolutely could make even Excel clone if you really want to.

The titles you listed are ARPGs. See the replies of previous posters regarding "engine being made for specific type of games".

Another problem is that an ARPG will require higher budget. As things are seen close and personal, and these days people would expect everybody to talk, meaning you'd need voice actors and such.

 

For such, you need jobs and burst at least, to cut performance issues. With 1k entities, you can reduce to almost inoticable overhead. Using DOTS entries on top of tha is just extra boost. But 2k is literały nothing, where dots can easily handle complex logic of 100k entities.

I don't know how you want to handle such performance mater in unreal. Unless you start code on source level of the engine. Possibly will eat more time, than just making features on Unity.

And with DOTS, you could increase world's entities AI to 10k, with some little effort.

 

Unreal have a ECS implementation called Mass. Haven used it myself just pointing out alternatives do exist

 

Thats very true, as Max have mentioned, Unreal have their own implementation but as far as i am concerned, its not as far as Unitys DOTS.

We will stay in Unity for now, lets see what results we can achieve, will defenetley post some updates on that

Thanks for all of your time

 

The next Witcher and Mass Effect games are being developed in UE5. Hard to tell if it's the power of Unreal Engine or their proprietary engines have become such messes that it's now cheaper to pay licensing costs to Epic instead of maintaining and developing their own undocumented spaghetti and training new staff to use something they've never seen before. Probably a little from column A and a little from column B. AAA is a different beast, though.

 

Yes, i have seen several reviews and talks.

As far as i learned is the UE Graphic better, and it doesnt consume as much compute capacity as in unity.

In terms of CPU, Unity seems to be a bit better (i dont know exactly why)

Examples compared were here base Models in each engine.

Quite interessting to me.

 

Yes, they are working on it, or has been released already.
Yet atm is just ECS.
Doesn't support multithreading, nor burtst type optimisation.
The primary gain is from bursting and multi threading. That can gain easily increase 10x performance. Then ECS can come on top of the cake.
I do wonder, how much Unreal ECS allows to boost the performance, from none ECS code.
Does anyone knows any benchmark? Are they planning multithreading blueprint support?

 

I don't know what machine code burst generates but I know il2cpp (another unity made compiler) is not as effective as .NET 5/6/7 JIT

 

Big name game in unity as example of what Unity can do is generally misleading, they have costly source access and even game like original has so much correction they are their own sub engine.