Download

Rate my upcoming pc build for UE4 and general game deveopment

Hello!

I am planning on upgrading to a new level.

This will be my first pc build. After some research I believe that these specs would be great suited for UE4 and other software’s such as Maya, substance painter, photoshop etc.

Link: https://pcpartpicker.com/list/3kDtZR

The build is managed to the highest budget.

My target is to create AAA content using the Unreal Engine 4. Shooter games such as battlefield v.

32gb ram is fine, you can always increase it in a year or so

single SSD drive 240gb is not enough. Get second ( m.2 ) for unreal projects storage. Best setup would be m.2 for system and 512 gb or more for unreal stuff. No worries here you can always get second SSD later. But at least get fast m.2 now.

And get 1080TI card. Let me explain this. Few years back i bought 980TI, my friend told me that I am stupid for wasting so much money. Since then he bought about 4 mid range cards to upgrade his pc. I am still at 980TI. Total money we spent on graphics cards is about equal now. But for last 5 years or so i enjoyed very powerful GPU, while his was always struggling. His latest purchase is finally at about equal level as my 980TI.

PS.
I think if you buy 1080ti card now it will be more than fine for next 2 console generations, ie about 10 years. Or until you want develop with raytracing card (but that is a gimmick for now, wait for gen 3 of those puppies).

[USER=“221053”]OBAN EMPIRE[/USER] I see you got a x470 mobo which is good since it has 4 memory slots, but you are using all of them at once, putting 4x 8GB RAM sticks, so in this case I would put 2x 16GB RAM sticks leaving 2 mem slots for future upgrades (there is a real gain here in terms of price of upgrade later).

I agree with @Nawrot about the SSD. You will be really good with 1 SSD for OS and 1 SSD for exclusive use for Unreal/other tools and working projects (M.2 NVMe from ADATA has great prices for 1TB - those are at least 10x faster than a regular SSD). For the 1 TB hard drive it is OK.

I can only confirm the suggestion to have a better GPU for now, if finding a 1080TI is still possible where you live, otherwise I would really wait for the AMD release of Navi 10, which performance is expected to match RTX 2080/GTX 1080TI and probably will cost less than both.

Thank you very much @Nawrot and @NilsonLima for the constructive feedback, as well as explaning the present state of the hardware timeline.

I will consider buyng 2x16gb ram so i can save two slots for future. Yes i will now prioritice for one more ( m.2 ) ssd for unreal.

I will work on one tittle ploject that is going to have a scope of 2 years (Battle royale shooter.) so that is the intention of this PC. That is why i wonder if i can trust this current 1070 card for this 2 years period? However, is it good to have the Ryzen 7 or should i downgrade to Ryzen 5 and overclock it?

I will most likly shovel in more resorces to the budget now. Bye bye paycheck :cool:

I think it is ok a Ryzen 5 2600x, cheaper and still enough threads for compiling shaders and stuff. I would just not going low on threads because time on compiling sometimes sucks, even more working on a big project. Also, even for hot environments, forget about overclocking and use the box cooler which comes with Ryzen 5 2600x if thats your choice, reports says that at stock it doesn’t heat too much and boost speed on cores very nicely. So, you are saving on another 2 components that you can use for the M.2 SSD.

1070 should get you trough to next (mature) generation of cards, it should be fine for 5 or so years. Also Ryzen should be just fine.

ps. If you have old PC you can always use it for distributed lightmass calculations, it speeds up time. But what really speeds up time is not creating huge and complicated level before you have game mechanics working. I know making levels is fun and easier thn doing rest of game, it is so tempting to bload source files with map that will be scrapped later.

Thanks again guys for the information! (@NilsonLima, @Nawrot) Iol i was just about to start making the mega level of the game. That is why i wanted a powerfull pc so i could start building the sick level in a faster phase. However, can i still make the level at first hand on a separate project and then merge it to the game project once it needs the level? or should i focus on making the game mechanics first and the level at last priority?

No matter of the type of the project: constructing a building, software, game, bridge, experience says you have few choices to avoid deadlines: build the most difficult things first (because at the start of the project you have plenty of time), then build the things that requires manpower (because you still have time), leave the things that can be quickly build and/or easy to integrate for later, when the time left for the project it not that much.

The approach above quickly deal with unforeseen issues, but requires that you have planned in advance the project and know all your tools and components, ie: game projects you need to understand the release for the engine strengths and its bugs to not get caught out of guard, acquired plugins need to be tested and known in advance, freeze upgrade on components that are already tested, because you can’t afford extra time to recover lost work later.

Big project teams can have different people working on different parts of same level. This is accomplished with level streaming where you have one map with landscape, another with illumination, another with props, and so on… they can all by loaded at same time composing a single level for playing. Check for perforce which is a versioning software and one of the best to work with game projects.

For years i was making levels for UT1999, UT2004, UT3 and so on. Thing with levels is that it must match gameplay mechanics. You need to have that mechanics in place and working to make level around it.

Very simple example: you have climbing or you do not. Lets assume you want climbing, so you make your level with that in mind, but then you realize climbing is not something that makes your game better. Result is that you need to rework whole level.

I would suggest you to get that PC you want and create small level, or two. Like one for raw testing of mechanics, like speed of walking running, jumping, if animation matches speed etc. And another one where you experiment with style of assets for your game, again small level. In that level you focus on quality of it.

I know making levels is fun, but do vertical slice of game features first. And reserve making of that cool level for days when you feel that derp had you cannot code game that day.

Great to know. I have already planned the project with raw testing levels and basic core mechanics for 1 year now (on my scrappy gaming notebook.) So my project has a more traditional approach with a detailed plan (now ready) then working full agile scrum. However, things can ofc be changed during time laps. The project biggest time taking task is actually the big main level, so that will be the first milestone execution.