• Godot Help

  • Recomended hardware for large projects and members of team.

Hi all. I'm learning how to write histories and I researching the price of a small game development company, but, I don't understand too much of GODOT or any other game engine.

So, I like to ask for those who already have some experience with programing and compiling what's a "good hardware" for test and compile a large game, about 816 GB files size in total with a large open world scenario like Night City in Cyberpunk or The Withcer 3.

In my case I'm gonna use a Linux based system, like Linux Mint or Arch Linux or Debian / Ubuntu and compile the game on Windows using a VM in the programmer computer or a Proxmox setup in another machine that will be used for quality assurance tests.

My intention is develop a MMO with a PS4 like graphics, like Killzone Shadow Fall.

And what it's a good team in your opinion, like 2 programmers or 5 or another number?

    RageousRaccoon with a large open world scenario like Night City in Cyberpunk or The Withcer 3.
    …
    And what it's a good team in your opinion, like 2 programmers or 5 or another number?

    CD Projekt RED

    the studio is formed by over 1000 talented professionals hailing from all parts of the world.

        Toxe

        A good amount with Blender and sculpting. So biological stuff and hard surfaces are fine, including textures and etc, no idea how to work with materials inside Godot since I don't read the manuals and etc. Also need to learn ArmorPaint for biological texturing and find a way to generate PBR from Blender (never tried).

        In programming with game engines a beautiful 0% skill level, I've learned the most basic programming stuff on a PHP course long ago and tiny bit of Pascal with Lazarus IDE (by myself), so I understand the concepts of objects, procedures, functions, compiling, libraries and little things like it, but not enough to know what a "good pc hardware" is. (compile ArmorPaint some weeks ago was a nightmare, but I got it).

        Tomcat Megalomaniak

        Over 1000 people doesn't mean people that actually works with the game development tiself, 2D artists, 3D artists, programmers, people that work with motion capture, marketing, social media, resellers, voice recording, musicians, translators, cleaners and much more, so the answer is to "open wide" and include too much things. And off-course, they have more than 20 years of existence.

        For a "start company", some people in Brazil develop Fobis St. Difna Hotel with a team of 510 people (closer friends), but they used Unreal with a skilled programmer (C++ if I remember right), but I don't have access to an programmer to ask directly.

        And off-course, Blender have some really nice features that can downgrade the number of people in a project, considering Geo-Nodes can generate extremely complex shapes and scenarios, so understand it will reduce the number of people in project, considering that large scenarios can be done in little time by one person and we can use things like Blender Kit procedural materials, that can be baked to regular textures and use tiling to optimize development of big textures like sand, dry earth, grass.

        We also have "4D AI" (ai motion capture based in light distance, even showed in a Blender conference).

        So for now, this reduce the "start of a company" to some artists, some programmers, someone very skilled to manage the server and motion capture, since audio can be a little generic in VFX in beginning we can just buy a package or hire some external musician to do it, same with OST.

        Considering that one person (that would be me) can create almost all scenarios and textures, and another person with more experience will take care of the skeleton and similar stuff.

        So I need to understand a little bit if Godot can support more than one programmer in the a project at same time, the amount of programmers that I've asked is to give me a idea of how much money I'll need based on value per hour of each one.

        So, the question is the same, what is a good PC config for programing in Godot?

        What is a good >start< number of programmers?

                RageousRaccoon Over 1000 people doesn't mean people that actually works with the game development tiself, 2D artists, 3D artists, programmers, people that work with motion capture, marketing, social media, resellers, voice recording, musicians, translators, cleaners and much more, so the answer is to "open wide" and include too much things. And off-course, they have more than 20 years of existence.

                reason I linked the WoW credits was precisely because it listed people and the positions they were in. I'd at minimum consider 1 person for each position in the US/home office. I think it's initially safe to ignore the other/international branches. But in terms of approximate amount of people needed I'd say it gives a good general idea.

                Once you have the key positions filled in, it's the people working the positions who can tell you how overwhelming their respective workloads are and how desperate for additional hands on deck they might be.

                    You haven't said how you're going to build the back end. If you're using godot, I don't think compiling is going to be a serious bottleneck. The bulk of the engine is already compiled (or you can compile it once). Gdextensions don't require recompiling the engine. Your computers should mostly be focused on the hardware necessary for playing the game and graphics design.

                    If you're building a server from scratch, than you're asking the wrong crowd.

                    And bear in mind that the more your developers are focused on windows, the better off you'll be. Your customers are using windows. Even for the back-end development, you can use windows. It's generally easier to work in virtual linux and ssh than virtual windows.

                      duane If you're building a server from scratch, than you're asking the wrong crowd.

                      And the wrong questions.

                        RageousRaccoon In programming with game engines a beautiful 0% skill level, I've learned the most basic programming stuff on a PHP course long ago and tiny bit of Pascal with Lazarus IDE (by myself), so I understand the concepts of objects, procedures, functions, compiling, libraries and little things like it

                        Then I would humbly suggest that you start small with your first game. Like, really small. Tetris, Pong, Breakout, Super Mario Bros.

                        Do not try to make an MMO as your first game. Anything related to multiplayer is pretty much the most complicated thing you can do. Do small and simple games first, especially before you take any amount of money into your hand.

                            RageousRaccoon My intention is develop a MMO with a PS4 like graphics, like Killzone Shadow Fall

                            You'll need at least 25 coders and 25 visual artists, and as many computers, excluding servers... given that you all know what you're doing.

                            On the second thought, more like 50 coders and 50 visual artists.

                                xyz You'll need at least 25 coders and 25 visual artists, and as many computers, excluding servers... given that you all know what you're doing.

                                On the second thought, more like 50 coders and 50 visual artists.

                                Without a basic semblance to a known budget, a design doc or a prototype game there is no way to estimate the manpower needed. These are just arbitrary numbers, as I'm sure is the point you are in a way making.

                                Tomcat Approximately such computers.

                                Again, arbitrary. Though I'd say based on OP mentioning PS4 games such as KZ Shadowfall as a reference I'd wager a HW comparable to a PS4 devkit might be the bare minimum.

                                • xyz replied to this.

                                      Megalomaniak Yeah I was a bit facetious here but given the question framing, I don't see how else would one answer. I almost suspect the OP is trolling. If not, here's more adequate answer - for producing a game on par with said titles you'd need a horde of people, at the least. And they'd all need to be highly skilled. How big should the horde be? Hard to tell without specific parameters. But rest assured it needs to be a horde 🙂

                                      @RageousRaccoon, game production is extremely demanding process, even for small games, let alone AAA-like stuff, not to mention MMOs on top of that. If you're really interested in doing it, start with internship (or a paid position if you can) in a studio that produces titles comparable to what you aim for, to experience firsthand how the sausage is made and what is the scope of involved resources.

                                          xyz I almost suspect the OP is trolling

                                          I don't suspect trolling, I think it's more like what is pretty much a cliché at this point: the beginner with no experience who cannot program and hasn't made even a simple Pong clone but wants to make an MMO as their first project.

                                          A tale as old as time.

                                              Toxe I think it's more like what is pretty much a cliché at this point: the beginner with no experience who cannot program and hasn't made even a simple Pong clone but wants to make an MMO as their first project.

                                              I thought this kind of ambition went out of fashion years ago. Apparently not 🙂

                                                  Well, I don't think ambition is actually a bad thing tho. Even if the first try fails, it probably teaches a lot and if it doesn't fail then hey, maybe you don't quite achieve what you imagined at the start but you still got something productive out of it.

                                                  xyz I thought this kind of ambition went out of fashion years ago. Apparently not

                                                  History goes round and round.

                                                    RageousRaccoon

                                                    I think the only one that can answer that is you. It's impossible for someone else to give an answer to that, as only you know what kind of game you want to make. And the resources needed depend on your scope and ambitions.

                                                    But here's some advice on how you can figure it out.

                                                    The first step you need is to define your project with a Game design document. Write down all the game mechanics, environments, enemies, players, art etc.

                                                    Then, gather some art references that come close to what style you want your game to be in.

                                                    Now you have a rough idea of what the game is that you want to make. With these things, you can now figure out what kind of talent and technology you need to make your game. If you don't have any experience making games, try to find senior people within each field (like an artist and games programmer/architect, etc.) to help you break down each element of the game and list what's needed.

                                                    Now, you might have a rough idea of the types of resources and talent needed. But then you have to ask within what timeframe you want to deliver. The shorter the timeframe, the more people and resources you need. With fewer people and resources, you need more time. So basically, 1 generalist artist and 1 programmer might be able to power through, but it might take you 20 years for the game you want to make. Or you hire 100 artists and 30 programmers, and you might finish the game in 4-6 years. These numbers are entirely hypothetical and made up (and there's more than just art and code in a game), though, as, again, only you can make the estimate.

                                                    Also keep in mind, the bigger the scope, the more important pipelines, workflows, and tooling become and you might need to invest in that as well.

                                                    Not entirely sure if I interpreted your question correctly, but hope it helps.

                                                        Megalomaniak reason I linked the WoW credits was precisely because it listed people and the positions they were in. I'd at minimum consider 1 person for each position in the US/home office. I think it's initially safe to ignore the other/international branches. But in terms of approximate amount of people needed I'd say it gives a good general idea.

                                                        Once you have the key positions filled in, it's the people working the positions who can tell you how overwhelming their respective workloads are and how desperate for additional hands on deck they might be.

                                                        I didn't see things in this way, so my apologies.

                                                        duane You haven't said how you're going to build the back end

                                                        What is a back end?

                                                        duane Your computers should mostly be focused on the hardware necessary for playing the game and graphics design.

                                                        So, if the recommended hardware for mediumhigh settings are Ryzen 5 3200, 16 GB of RAM and a 1060, I can use pretty much the same hardware for programming, or adjust a little bit more for a Ryzen 7, 32 GB RAM and a 2060 or AMD equivalent?

                                                        The "starter" workstation for Blender sculpting and Armor Paint it's something like Ryzen 9 3D with 128 GB of RAM and a 24 GB RAM GPU with NVME (two disks in raid 0 if possible), the ideal would be a Threadreaper 3264 cores with 1 TB RAM for large procedural materials that will be exported to textures, but I think it's too much for the programming.

                                                        duane If you're using godot, I don't think compiling is going to be a serious bottleneck.

                                                        Now we almost hit a answer, I'm trying to figure out what it's important in the hardware, let's say that a single sata 3 sdd 450 mb write speed will compile the game in 10 minutes, so 1 NVME 2000 can decrease this time to 2:30 if processor and ram are fast enough to write all this data, the balance between processor - ram - hard disk is my main doubt, I don't see to much sense in buy a Threadreaper with 64 cores and 2 TB of ECC DDR 5 for a single NVME 2000 mb, like I don't see any sense in get a single XEON with 64 GB RAM and 20 NVME over 7000 mb in raid 0 , I need to find the balance between things, but I'm not a programmer so I don't what is "most" important, and what is "just" important.

                                                        For sculpting in Blender for example, I know that processor and RAM are the "most" important and GPU is "just" important, a 2060 will be fine. (hard to believe but's true).

                                                        For texturing in other hand, half of RAM and processor of the sculpting machine will be enough, but at least a 3090 (prefer 4090) it's recommended.

                                                        For godot and programming I have no idea.

                                                        MikeCL But here's some advice on how you can figure it out.

                                                        MikeCL try to find senior people

                                                        That's why I'm here, trying to look for seniors in Godot to figure out what is important in hardware for programming. Tons of TB of hard disk, tons of GB of RAM, if a Thredreaper is required or a Ryzen 7 do the job.

                                                        MikeCL you might have a rough idea of the types of resources and talent needed

                                                        For art yes, for programming no.

                                                        MikeCL Not entirely sure if I interpreted your question correctly, but hope it helps.

                                                        It helped a lot. My problem is find a programmer that already worked with games and can explain his difficulties, argument why the X point of a project is more complicated than the B point and why Y tech is a good or a bad idea for each kind of game.

                                                        It's possible to rent a server like machine with any Linux based distribution and test the database and speed and etc. So a server it's not important for now, specially considering that It's possible to run PROXMOX in almost any machine and do the very basic connections and security tests.

                                                        duane And bear in mind that the more your developers are focused on windows, the better off you'll be. Your customers are using windows. Even for the back-end development, you can use windows. It's generally easier to work in virtual linux and ssh than virtual windows.

                                                        Windows is around 1600 BRL (Brazilian money), Adobe around 2400 BRL a year, ZBrush around 2400 BRL a year and a minimum payment for 20 days of work 8 hours each day is around 1200 BRL (7.5 BRL / Hour), just ONE machine will cost 6400 BRL a year, with 20 machines (128.000 BRL), for 30 machines (192.000 BLR), I can hire a little group of very skilled designers or programmers for two years or more.

                                                        My intention is have a company that develop Linux games and them port them to Windows with 0 software cost, so I can have better machines and better people around, Mac will be out of the list because a single mac mini is around 7000+ BRL with minimum config.

                                                        Windows will be only for compile the project and run tests on a test machine.

                                                        Toxe Then I would humbly suggest that you start small with your first game. Like, really small. Tetris, Pong, Breakout, Super Mario Bros.

                                                        Do not try to make an MMO as your first game. Anything related to multiplayer is pretty much the most complicated thing you can do. Do small and simple games first, especially before you take any amount of money into your hand.

                                                        I quote the MMO, but it's a long time project to be developed, and it was a bad idea to comment about it, I fully understand how hard it is to record all mocaps, find out people that is able to do mocap, create game styles for each character, bones for body and clothes, a create your character with some customization (at least with models like Hogwarts Legacy or Cyber Punk), database design is too complex and etc, considering all medium works like edit textures and etc, so I'm gonna release smaller 3D projects, like a 35 hours game or so.

                                                        Maybe something like Kena (a simplified version of it), easy scenarios for Geo-Nodes, possible to repeat textures a lot of times and 510 enemies, and many mocaps can be found in free packages, Adobe have a web site with a lot of them.

                                                        xyz You'll need at least 25 coders and 25 visual artists, and as many computers, excluding servers... given that you all know what you're doing.

                                                        On the second thought, more like 50 coders and 50 visual artists.

                                                        That's a solid answer, so know I know I can start with less than 25, and the indie games can be used to hire more people over time, until reach 50 and later 100, and so on.

                                                        xyz game production is extremely demanding process, even for small games, let alone AAA-like stuff, not to mention MMOs on top of that. If you're really interested in doing it, start with internship (or a paid position if you can) in a studio that produces titles comparable to what you aim for, to experience firsthand how the sausage is made and what is the scope of involved resources.

                                                        Brazil have one game development company PULSATRIX and get a job in there right now it's not possible, due to some healthy problems I never created a 3D portfolio, I have a lot of knowledge but nothing showing it, so one step at time. I'm just trying to reach the PC price for programming and other stuff to start a crowdfunding campaign to it.

                                                        And like I've said, Geo-Nodes and Blender Kit reduces 80% some times more or less depending of what is being done and how is being done, so I can literally develop the entire "world" of a game in a couple hours, for people with extreme knowledge about it, it's even possible to develop the entire base (sketch) 3D of Night City in a couple of hours and them create some builds and assign to the blocks, there is even thing like city generators for Blender, with texture, baking and little adjustments, it's possible to build it in less than a month with one really skilled person or 2. Off-course, repeating shapes and textures like NC does and no internal apartments or furniture.

                                                        Programming is a real issue because there is no shortcuts like Geo-Nodes and some other decisions like use GPU Draw inside of Godot for grass instead of regular meshes for blender and hit-boxes for attacks, particles, dynamic shaders, hit-boxes for body when character is hited.

                                                        Toxe I don't suspect trolling, I think it's more like what is pretty much a cliché at this point: the beginner with no experience who cannot program and hasn't made even a simple Pong clone but wants to make an MMO as their first project.

                                                        The MMO its a long time project, not first one. I'm ambitious not a complete idiot.

                                                                                    RageousRaccoon How about this. For start you team up with one (1) coder, and together you develop a very polished 5-level 3D Zaxxon reboot, sauced with some character based storytelling inbetween levels and never before seen visual design, also adding a few innovative mechanics to the base game. The coder dude does all the coding and you handle the rest. Use computers you already have. See how much or little effort/time/computing it will take. Might help with getting some perspective for future bigger projects.