CPU & @GHz for 60+FPS

[Drzator]

ATM i have R5 1600 @ 3.9GHz, proper GPU, 16GB RAM will be soon here. Somehow i doubt that this CPU will be enought for rock stable 60FPS.
Obviously i have 2 options, weiting for ryzen 3000 or just to buy i5 9600k.

Someon here whit OC coffe lake ~5GHz?

[Kadatherion]

I have a rather abysmal i5 4460, 3,2ghz, with an only decent 6gb gtx 1060, and I do have somewhat stable 60fps at 1080p with settings inebtween high and ultra (basically, all on ultra except for the LOD, that's on high equivalent settings). Fps go lower very close to some stations (inside them the hit is for everyone, station interiors with still everything going on outside are badly optimized for everyone), but for 80% of the common fly around, do things day to day gameplay it's been alright. So I'd say you should have an overall decent experience.

Performance tanks for me however at 1440p, even lowering the graphic settings, which is a bit disappointing, I was hoping I could play at 1440 decently by maybe just toning down a few settings, but that gives me an average fps more in the range of the 40s.

[Tiliam]

9900k at 5.1ghz OC across all cores here. Other components: 1080ti with mild (+8%) OC, 32GB of 3200mhz CAS16 RAM, modern SSD.

Getting between 70 and 120 fps depending on where I am at 1440p with everything on ultra/high (SSAO on, glow on). Have SS set to 4x MSAA.

[PabloRSA]

The game is now Vulkan the bottleneck will now be your GPU as it moves alot of the rendering to the GPU and not the CPU.

My specs
CPU: Core i7-4790K @ 4.0Ghz | Motherboard: Asus ROG Maximus VII HERO | CPU Cooler: Xigmatek Dark Knight (Night Hawk) |
| GPU: EVGA GTX 980 Ti Superclocked+ ACX 2.0+ | RAM: 32GB Vengeance Pro DDR3 @ 1866MHz C9 | SSD: Samsung 850 Pro 256GB ( Windows 10) | SSD: Intel 120GB |HDD: 6TB

My CPU is at around 35%, and memory useage is around 30% in a busy system
I get a avarage 80-90fps in astroid systems otherwise usually 120+ fps, this is on High shadows/detail, 4x MSAA (max setting available) with glow and SSAO turned off with these last two options on I drop to around 45-60 but CPU never changes.
Game installed on Samsung SD, this does help with load speed by alot compared to Game HDD..

Hope this helps you.

Edit: forgot to metion im using a ultrawide 2560x1080

[Drzator]

Tx all for info.

9900k at 5.1ghz OC across all cores here. Other components: 1080ti with mild (+8%) OC, 32GB of 3200mhz CAS16 RAM, modern SSD.

Getting between 70 and 120 fps depending on where I am at 1440p with everything on ultra/high (SSAO on, glow on). Have SS set to 4x MSAA.

Can you do 1080p low setting test plz?

Looks like ryzen CPUs are close to enought, someon here whit R5 2600 @ ~4.0GHz ?

[sim787]

I have a Ryzen 1700X and a 1080 Ti, get a solid 60fps in 1440p, it does actually run quite well in 4k but there's big drops in the stations

[Drzator]

OK, i will try whit current CPU, alternative is R5 2600x, we will see FPS. Better late on patry then Vsync OFF .

[Baconnaise]

I'm primarily using a i7 4790k oc'ed to 4.8. GPU is a R9 Fury. Playing at 1440p with most of everything bumped or maxed I think. Docks are the only place where it can bog down. Runs like butta otherwise. AMD cards typically do well with vulkan since you know...

[Tiliam]

Tx all for info.

9900k at 5.1ghz OC across all cores here. Other components: 1080ti with mild (+8%) OC, 32GB of 3200mhz CAS16 RAM, modern SSD.

Getting between 70 and 120 fps depending on where I am at 1440p with everything on ultra/high (SSAO on, glow on). Have SS set to 4x MSAA.

Can you do 1080p low setting test plz?

1080p low capped at 144fps (can't get vsync off atm). GPU usage dropped from 99% to 23% so the frames would be a lot higher if I had vsync off.

[Ranix]

Ryzen 1920 here at 4ghz with 3200mhz RAM and a 980ti, enjoying reasonable framerates. I am usually locked at 60fps (vsync) but will sometimes drop down to the high 30's on station landing platforms in high-traffic sectors with the map open.

[CBJ]

The game is now Vulkan the bottleneck will now be your GPU as it moves alot of the rendering to the GPU and not the CPU.

Sorry, but as has been explained before, this is quite simply incorrect. More of the rendering load may have moved to the GPU, but the processing of the game universe is still on the CPU, and unless you choose graphics settings that are too high for your hardware, it's the universe processing that will be the limiting factor.

That is also the reason why a "rock stable 60FPS" is not an achievable goal in this kind of game. It will always be possible to find or create situations in which the game slows down, simply by having enough going on nearby while the game still has to simulate the rest of the game universe that is further away.

[PabloRSA]

The game is now Vulkan the bottleneck will now be your GPU as it moves alot of the rendering to the GPU and not the CPU.

Sorry, but as has been explained before, this is quite simply incorrect. More of the rendering load may have moved to the GPU, but the processing of the game universe is still on the CPU, and unless you choose graphics settings that are too high for your hardware, it's the universe processing that will be the limiting factor.

That is also the reason why a "rock stable 60FPS" is not an achievable goal in this kind of game. It will always be possible to find or create situations in which the game slows down, simply by having enough going on nearby while the game still has to simulate the rest of the game universe that is further away.

Im not a game dev and new to vulkan, just recently reading about it and still trying to understand it, but form what im reading OPENGL does alot of rendering/debug/etc.. tasks on the CPU before moving to GPU and vulkan does not, it now sends everything to the GPU by pipelines for you to do it without any helpers/wrappers (still dont understand most of it yet)
Thats what i mean by shifting more of the load of rendering to GPU it uses less CPU power. OFC game ticks/update can only be done by the CPU.
Which in theory should make the GPU more of an issue/bottleneck, EG like the tech forum posts, VRAM needs 3GB min which has created a bottleneck for the low end systems, thats what i meant by that statement.

[EmperorDragon]

Well, I tested the game on a 1.5GB GTX660, and it runs! Actually, I got a solid 40 FPS but drops to the low 20s when flying through nebulae or during large battles. FPS goes up even more if I turn down the graphics settings.

Of course, crashes and UI glitches are a given with half the recommended video memory but, the fact that the game itself runs quite smoothly on that card is... interesting.

As for CPU, I run an i7 4770 and it never breaks a sweat, not yet at least. I don't have a lower-spec CPU to test but I reckon anything on that level or higher along with a 4GB GDDR5 card (maybe even 3GB?) will give pretty good framerates most of the time if you're willing to reduce graphics settings a tad bit.

[KextV8]

I'm running i7 2600k OC to 4.5ghz 16gb RAM, OC GTX 970.
FPS 100-110 on highways.
FPS in 80-90 in normal play.
FPS in the 60s in really graphic crushing areas like the heavily particle filled Grand Exchange Asteroid areas.

Resolution 1080p.
Graphics settings:

Spoiler

Show

Game stays around 30-40% CPU utilization with jumps into the 50s if I teleport across the universe as it loads the new area, and maxed out GPU utilization. I figure the instant you lock vsync off, the game is GPU bound unless your GPU is ridiculously stronger than your CPU. I'd probably have to have 1080 TI or RTX 2080 before it could process graphics faster than my 8 year old CPU can feed it data. And that's at 1080p. At 2k or 4k I'm not even sure you could CPU bind the game unless you created a ridiculously heavy load, like thousands of player owned objects. If you do manage to CPU bind the game, you're probably going to end up slide showing. Some people will definitely manage to do that, but I probably never will. I'm not into the build portion of the game really.

Biggest advice I can give to anyone having FPS issues is to ditch MSAA. I cant really tell the difference visually between that and FXAA, but it seemed to cut my FPS in half when I had it on. The other main thing is whether your card has enough ram. If your GPU has a low amount of RAM, the single best setting you can lower is texture quality.

[Socratatus]

I have a ryzen 2700X, 16 gig ram, but just a lowly 1060 6 gig. Games runs fine running on Ultra at 1080. I haven`t checked FPS because I`m happy. I don`t normally check Framerates unless I notice issues.

p.s. Also on an nvme m.2 SSD.

[Ranix]

I have a ryzen 2700X, 16 gig ram, but just a lowly 1060 6 gig. Games runs fine running on Ultra at 1080. I haven`t checked FPS because I`m happy. I don`t normally check Framerates unless I notice issues.

p.s. Also on an nvme m.2 SSD.

hey I was thinking about getting one of these, how are your save times? For me, on a standard sata ssd, it takes about 40 seconds to save a game

[CBJ]

Im not a game dev and new to vulkan, just recently reading about it and still trying to understand it, but form what im reading OPENGL does alot of rendering/debug/etc.. tasks on the CPU before moving to GPU and vulkan does not, it now sends everything to the GPU by pipelines for you to do it without any helpers/wrappers (still dont understand most of it yet)
Thats what i mean by shifting more of the load of rendering to GPU it uses less CPU power. OFC game ticks/update can only be done by the CPU.
Which in theory should make the GPU more of an issue/bottleneck, EG like the tech forum posts, VRAM needs 3GB min which has created a bottleneck for the low end systems, thats what i meant by that statement.

You've understood the basic principle of Vulkan, but have not drawn the right conclusions from it. You are only reducing the amount of processing required for rendering. In a game multi-threaded game, rendering isn't necessarily the limiting factor in the first place. That is very much the case with the engine used in X4.

The simulation does run on a separate thread from the rendering, but you cannot decouple the two completely. It's no good the rendering thread saying "oh, I've finished one frame, I'll do the next one" when the simulation thread hasn't finished updating things yet. In the best case you'd end up rendering the same thing twice in consecutive frames (which is essentially the same as halving the framerate again), and in the worst case you'd end up with the simulation thread being half way through the job, and the render thread rendering an inconsistent state with some things updated and others not. In practice it's more complicated still; efficient threading requires sensible locking schemes that prevent nasty clashes (and crashes!) when one thread is trying to read while another tries to write, and there is some processing that has to happen on one thread which you might think could happen on the other in order to make the whole thing work. So in summary, no matter how fast the rendering is, it still has to synchronise with the simulation, and I'm afraid this means that your suggestion simply doesn't work.

Now before you come up with the bright idea that things you can't see could be updated in the simulation less often than those you can, we're way ahead of you there. In fact we've gone beyond that and managed to make it so that some things don't need to be updated at all, and will just be able to give the correct information whenever it is needed rather than needing frame-by-frame attention. Things like asteroid rotation, factory production, and many others fall into this category. Without all this the universe simulation wouldn't even be viable in a non-VR game. You might also think you could magically speed things up by running the simulation on multiple threads, but this is a lot harder than you might imagine. The objects in the game are very heavily inter-connected (consider a ship, trying to dock at another ship, which in turn is flying to a station but has been interrupted by an attack from a group of enemy ships that are all coordinating among themselves) which makes it next-to-impossible to divide it into discrete units which can be processed independently. Quite apart from the complexity, you can quickly run into threading overheads, and that's without even considering the high chance of it all grinding to a halt with one object waiting for a lock on another to be freed, which in turn is waiting for a lock on another object, which itself is waiting for a lock on the first one!

Believe me, we have put a lot of time and effort into analysing this and have explored many, many different possibilities. While we will always be looking for improvements and optimisations, we are quite sure that there is no simple solution which will suddenly make it possible to double the framerate!

[Socratatus]

I have a ryzen 2700X, 16 gig ram, but just a lowly 1060 6 gig. Games runs fine running on Ultra at 1080. I haven`t checked FPS because I`m happy. I don`t normally check Framerates unless I notice issues.

p.s. Also on an nvme m.2 SSD.

hey I was thinking about getting one of these, how are your save times? For me, on a standard sata ssd, it takes about 40 seconds to save a game

It takes about 20 seconds. Just timed it and it took 18.38 seconds while docked at a starbase. This is also while running a video I`m watching on my other monitor.

[MrMikey]

Getting very stable fps of 98-100 (vsync on with 100hz screen) in 3440x1440 with very consistent frame timings even later in the game with loads of action happening but the specs are quite overboard but its really good to see the game take advantage of it (9900k slightly oc'd+16gb 4000 mem with good timings,2080ti's (not using sli in this title))

[Scoob]

Just so everyone is aware, most tools for measuring CPU load don't really paint an accurate picture when HT is enabled.

With my 2600k @ 4.6ghz it looked like my CPU was averaging about 30% load over all threads, with three threads a fair bit busier than the others, one particularly heavy. So, my CPU, though somewhat more busy on a couple of threads, had ample capacity to spare, so it couldn't be a bottleneck, right? However, this is more down to the tool thinking it's got two full Cores to play with, not one core with some clever scheduling tech. If you temporarily disable HT, you'll see a much more accurate representation of how hard X4 pushes a CPU. for me, it's nearer 70% than 30% which is a huge difference.

HT/SMT can be great, allowing a single core to run dissimilar loads concurrently, or taking advantage of available processing power when an already running process is waiting for something - be it the result from another process, memory access etc. This results in pushing the actual single core harder or at least keeping it productive. A good tech. However, it really does seem to confuse monitoring tools. I personally use Process Explorer, it's pretty solid but misrepresents things similarly to Task Manager it seems.

It's often more realistic to ADD the load of threads 0 & 1, 2 & 3 etc. together to get an idea of the actual load on the underlying single Core. This might seem strange as, when stress testing for example, we see ALL threads 100% loaded. However, these are very specific workloads that are getting every single ounce of performance out of the CPU, using HT/SMT to its fullest. Game workloads rarely work this way. I think the only game I've tried which loaded ALL thread really to their limit was Crysis 2, that game really could leverage every bit of CPU power available.

I just wanted to mention this briefly, and not going into much detail, as, initially, I too though my CPU really wasn't being pushed as hard as it was. The game, for me, appears to run just as well with HT off as it does with it on. However, with it off, my CPU looks like it's working so much harder, due to how Process Explorer sees the work load and available threads.

All that said , I do generally get 60fps at custom setting based off the Ultra preset, with Glow=Medium and 2xMSAA @ 1920x1200 on a GTX 1070 running at 2ghz. Of course the game dips below this quite often, but not hugely. The biggest drops I see are while in the Map screen, where FPS isn't quite so important. During flight, combat etc. it's fine.

I do tend to let my pilots fly me around a lot of the time, while I enjoy the external view. Game is generally running really quite well for me, so much so I've indefinitely postponed any PC upgrade plans - I had a 9900k as associated bit sat in my eTailer basket lol.

Scoob.