D.Va

I have 16GB RAM.

The most important thing on a motherboard is how many USB connections it supports, since that sticks with you for many years to come. You can expect a computer to last 10 years, but the graphics card will need upgrading every 3-5 years if you want to stay on top. That's why I advocate only buying the low and mid-range graphics cards. It's better to get a weaker one and upgrade more frequently. I bought my PC in 2013 and upgraded the graphics card in 2018 (which I could have done as early as 2016, but I wasn't playing PC a lot at the time) and expect to upgrade the graphics card again in another 2-5 years. Graphics card get old a lot faster than processors. Even the i7-2600K from 2011 still performs 80% as good as the newest processors. In my day, K's came with stock coolers and spending additional money on 3rd party coolers was always cost inefficient. Apparently, K's don't come with stocks anymore, which may or may not change the cost-efficiency situation. Computers make do with memory. I can run one game and use 9GB, but someone who only has 8GB runs it and only uses 8GB without adverse side-effects.

If anyone here has an iPhone, you can adjust your monitor brightness with it, like this: Turn down your iPhone volume to zero. Click the volume up button four times, which puts volume at 25%. Open the control panel or whatever it's called and increase brightness to 25% so it matches the volume bar. Hold your iPhone in front of your monitor and adjust the brightness on your monitor until the brightness matches. You may be surprised at how dim this is, but this is in fact the recommended brightness for office environments all over the world (300 candela, if I remember correctly) and it greatly decreases the amount of headaches you get after prolonged use. You can also adjust colour this way, but iPhone whites are possibly slightly blue. With this brightness, I find Gamma 2.0 looks the most natural.

I've redesigned my guide to be a hell of a lot more aerodynamic than other guides, but I'm pointing to your guide (and soon other guides) for more detail in my guide. Comments on your recommendations: Civ. Traffic doesn’t affect performance by more than 1% my any of my benchmarks to date. Shadows HIGH compared with MEDIUM makes a 0% performance difference to me. I’ve yet to test Res. of Cockpit Displays by viewing a lot of action through the Warthog pod or whatever may stress this option. MSAA isn’t necessary in 4K I guess? Cool! Lens Effects are personal preference, but don’t affect performance by more than 1% in my tests. I have a professionally calibrated monitor and Gamma 2.0 looks a lot better than 2.2, but this will vary depending on your monitor brightness. Other than that we’ve come to a lot of the same conclusions it appears

After discovering a few optimisations that made the whole MSAA/Shadows/Visib. Range choice rather redundant, I've re-written the entire guide and made it an easier choice of Low, Medium, or High settings instead!

I did originally have an "intermediate" recommendation, which gave +55% compared with minimum options and -55% compared with maximum options. I may bring that back at some point. Your options and performance are already quite optimised! What do you want? Try this: Textures HIGH Terrain Textures HIGH Civ. Traffic HIGH (almost free) Water HIGH (almost free) Visib. Range ULTRA (You can also try decreasing this to HIGH) Heat Blur HIGH (doesn't make any difference to me, but it may affect your machine) Shadows LOW (Set this to FLAT ONLY for better performance, but worse cockpit shadows) Resolution PERSONAL PREFERENCE Aspect Ratio PERSONAL PREFERENCE Monitors PERSONAL PREFERENCE Res. of Cockpit Displays 1024 EVERY FRAME MSAA X2 (You can also try increasing this to X4 or X8, if Visib. Range is set HIGH) Depth of Field OFF Lens Effects DIRT+FLARE (probably free) HDR OFF Deferred Shading ON Clutter/Grass 1000 Trees Visibility 80% (You can change this in-game without exiting a scenario and see if turning it up/down makes any practical performance difference) Preload Radius 60000 Chimney Smoke Density 0 Gamma 2 (You can play around with this in-game also) Anisotropic Filtering X2 (probably 5% performance increase) Terrain Object Shadows FLAT (unless it kills your performance) Global Cockpit Illumination OFF Disable Aero Interface OFF Vsync OFF Full Screen ON Scale GUI OFF

It might have chaned in DCS 2.5? I also never noticed it, but when flying at medium altitude in Caucasus, it made the difference between shadows only being rendered almost just underneath me, or shadows being rendered pretty far away.

The system requirements aren't insane at all. You can probably almost max out all options in 1920x1080 playing with any gaming CPU released since 2010 and a low-range graphics card from 2016 (the 1060), a mid-range graphics card from 2014 (the 970), or a high-end graphics card from 2013 (the 780)... and it looks awesome too! I'm playing it with a 2013 CPU and a 1060 (the cheaper 3GB VRAM version, even). You shouldn't have performance problems, if you didn't before.

Yeah, DCS 2.5 looks amazing and runs amazing. We have almost the same specs, so you can probably get a lot of use out of my optimisation guide: https://forums.eagle.ru/showthread.php?p=3381949

I already know everything you wrote. But, no matter what the theory is, I did in fact get 2.5%-5 performance differences with all Texture options at minimum instead of maximum, Water MEDIUM instead of HIGH, and a bigger 10%+ performance difference with Clutter/Grass minimum instead of maximum, and about the same difference with Cockpit Global Illumination. This was fully replicable, until I changed resolution. Also, the Texture options make 0% VRAM difference for me. Instead, VRAM rather randomly changes from 2.6 to 3GB (all I have) when restarting. I've been unable to find any option that affects VRAM consistently, though I haven't modified the big three yet: MSAA, Shadows, and Visib. Range. In other words, VRAM works quite unintuitively and is difficult to benchmark with my current methods. The difference between EXTREME and ULTRA is big. I benchmarked the difference around Batumi and the difference in range at which you can see buildings, vegetation, their density, and the distance at which shadows appear is alarming, in my opinion. MSAA X8 and MSAA X4 has no performance difference on my machine. If you play with a weaker graphics card or at a 2K-4K resolution, multi-monitor, or VR, your experience may vary. Wow, that's a lot of vegetation North-east of Sochi. I get 40-45 FPS there (more if the sky covers more than 1/3 of the screen), 38 FPS stable when I dive deep into the most heavily vegetated areas and in one valley I managed to bring it down to 37 for a few seconds. It also drops to a little less than that for a split-second when I initiate turns and when flying into the area I think it temporarily dropped into the low 30's for a while. My specs, again: 1920x1200 (+10% performance in 1920x1080) i7-4770K (4.0 GHz "OC") ASUS 1060 Dual (3GB VRAM) 2x8GB 1600 MHz RAM DCS installed on secondary Samsung 840 Evo 250GB Windows 10 64-bit Can't think of anything else that affects my performance. I'm currently not using my TrackIR 5. Mirrors are off, naturally. I flew the Mirage. I'm using the exact options I describe above. (The top-left says 39 FPS, but it was 38 FPS before I paused)

I don't play online, at all, that's true. Processors without stock coolers?! Madness! 8GB is more than sufficient for doing anything other than playing simulators, like ARMA and DCS. Windows uses 2GB and most games use 2-4GB. Widows+DCS 2.5 uses 7-9GB in my experience, but allegedly it uses a lot more if you play online. I agree about image quality over refresh rate.

I challenge you to prove any of my statements wrong - I stand by them. I don't know if I asked, but what resolution do you intend to play at? You wrote something about a 1080p monitor, a 1440p monitor, and another 1440p monitor you're going to get? I play DCS with a 1060 and I can play with almost all options completely maxed out and get 30-60 FPS (average 45) in 1200p with only one monitor... I've never played with multiple monitors, but I can't imagine it would run great without a 1070 minimum. I don't know where you live or what your stock situation is (American stocks are destroyed by the Bitcoin miners), but I'd be surprised if there wasn't anything like a "MSI Z270-A Pro" motherboard, or "Asus Prime Z370-P", or any other basic motherboard in the $120 low-range instead of the $210 high-range (which is what the Aorus Ultra Gaming costs to me) available anywhere. Regarding 8400 or 8600K, it's up to you. I'd go for the 8600K, because the upgrade is reasonably cost-efficient anyway. I would avoid VR currently. The current generation $600 VR headsets came out in 2016 and it appears the second generation is going to start releasing already in 2018, which will may bring greatly increased quality, increased performance, or increased cost-efficiency. We don't know which of the three, but we expect to experience it in 2018-2020. You also need AT LEAST the 1070 really. Regarding RAM, some people have advised me that DCS uses a ton of RAM in multiplayer. I don't know, because I don't play multiplayer, but based on that 16GB is certainly advisable.

I've figured out a few more optimisations that I'll work into the first post after I've benchmarked them in a wider variety of scenarios. Also, for users of cards with only 3GB RAM, I'm experimenting with ways to manage VRAM. If it runs out, there's a significant 15-25% decrease in performance, but I haven't been able to identify the options that affect it yet. Here are my current recommended options (1920x1200): Textures HIGH Terrain Textures HIGH Civ. Traffic HIGH Water HIGH Visib. Range EXTREME Heat Blur HIGH Shadows HIGH Resolution PERSONAL PREFERENCE Aspect Ratio PERSONAL PREFERENCE Monitors PERSONAL PREFERENCE Res. of Cockpit Displays 1024 EVERY FRAME MSAA X8 Depth of Field OFF Lens Effects DIRT+FLARE HDR OFF Deferred Shading ON Clutter/Grass 1500 Trees Visibility 100% Preload Radius 1500 (don't know if it makes any difference anyway) Chimney Smoke Density 0 Gamma 2 Anisotropic Filtering 4X (5%+ performance compared with 16X, no significant graphical difference)* Terrain Object Shadows FLAT (DEFAULT currently has extremely low resolution shadows and this actually looks better) Global Cockpit Illumination OFF (7.5%+ performance, no significant graphical difference) Disable Aero Interface OFF Vsync OFF Full Screen ON Scale GUI OFF I've discovered Anisotropic filtering and (especially) Global Cockpit Illumination affect FPS in practically significant amounts and Textures, Terrain Textures, Water, and Clutter/Grass APPEARED to affect performance, but in another test they affected nothing. After adjusting Anisotropic filtering and Global Cockpit Illumination, I get like 50 FPS average in 1080p and 40 FPS average in 1200p, which rocks. I've tested this in the Warthog, Mirage, and Huey across Vegas and Caucasus maps, high and low altitude and performance rocks across the board.

I've made a lot of benchmarking with regards to VRAM and yeah, when it runs out my FPS dips from about 40 to 30 (15-25% difference), but WHEN it runs out is completely random. Simply restarting the game randomly changes the amount of VRAM DCS uses between 2.6-3GB. I'll need a lot more testing for this. It sounds like we have about the same performance. Just to clarify: what FPS did you have if you started the scenario as I asked and didn't touch anything, just hit auto-pilot and waited like 10 seconds. About 45-50 fps? I also get dips down to about 30-35 when diving towards the vegetation north of Kobuleti airfield and stuff like that, so it's probably near-identical.

After benchmarking a lot with my 1060 3GB I've discovered: • Yes! 3GB vs 6GB VRAM may affect your DCS performance. • DCS on almost maximum options uses almost exactly 3GB VRAM in 1080/1200p, which means only a minor adjustment in settings can put you either above or below the limit of 3GB VRAM. • Playing with almost maximum options and Terrain Textures MEDIUM puts me at 2.8GB VRAM and Terrain Textures HIGH puts me at 3GB VRAM (all of it) and the difference is 38 versus 32 FPS, a significant difference! I don’t know if the difference is absolute or relative and I’m going to experiment with options that put me just below the limit of 3GB VRAM. With a little luck, I may be able to improve performance 10%+ across the board for players with 3GB VRAM with only a few minor and probably invisible adjustment of the options. • DCS uses a minimum of 1750MB VRAM on my machine in 1080p. • DCS maxes at 4GB or around there VRAM, based on what I've witnessed in other benchmarks. I’m now considering VRAM as an individual resource in my benchmarks and will benchmark how much VRAM important options eat.

Oooh, thanks!

But DCS appears to run in perfect full screen for me even with full screen OFF. Alt-tabbing doesn't affect anything, but I understand what you mean, so it may be different for different people. The problem with FXAA is that it blurs all text in the game, including the menues! Doesn't it for you? ARMA 3 has fantastic FXAA by the way.

What's wrong? Why isn't my beautiful avatar that is <1028x1028 and <1MB showing up? It's visible on my profile page, however.

I'm going to write a computer guide for everyone upgrading, but maybe not before this weekend. Anyway, you'll want something along the lines of: Case: personal preference. I recommend a quiet one! Motherboard: Any that matches your processor, don't overspend here! Processor: i5-8600K Memory: 8GB RAM (yeah, it's enough) Graphics card: as expensive as you can afford! Storage: Samsung 860 EVO 500GB Optical: if you want an optical drive (aka CD/DVD), get any Power supply unit: any that supports your CPU/GPU, don't overspend here! Additional cooling: none! Operating system: Windows 10 64-bit These are the 10 components of a computer. All of this, excluding graphics card, should go for less than £1000. In other words, your budget is a lot of money and I would spend that money on a 4K monitor (500€-ish) and 1080 or 1080 Ti card (500€-ish). Regarding monitors, you will have to choose 60 Hz with higher quality image or 120/144 Hz with lower quality image.

First of all, know that upgrading to an 8XXX CPU wouldn’t give you more than 20% better performance (10-30% depending on game) in gaming. Personally, if I was you, I would buy the i5-8400K or the 25% more expensive and 5-10% stronger i5-8600K and forget about overclocking (it’s never cost-efficient in my experience), go with stock! Water cooling is a cosmetic upgrade from air. If you forget about overclocking, you can simply get any decent brand motherboard that’s about $120 instead of the $180 options and save money. Also, you don’t actually need ANY RAM upgrade. DCS 2.5 has a memory leak problem (some say, I haven’t experienced it), but when that’s fixed you won’t need any more than 8GB RAM to run DCS. My DCS uses 9GB. But the thing is this: if you only have 8GB RAM, your computer makes do with what it’s got, and I don’t believe there will be any actual performance degradation in practice. The graphics card is indeed not upgradable without going up-stream against cost-efficiency; however, if you’re going to play with THREE MONITORS (I only use one!) then if you save your money on CPU ($75), motherboard ($60), water cooling ($100), and RAM ($250) as I’ve described above, you’ve already paid for two thirds of a 1080 upgrade. Have you considered going 4K or VR instead?

I watched the videos. The first one didn't tell me anything I didn't already know and the second wasn't relevant at all, was it? What exactly am I supposed to see that means I need more than 4GB VRAM here? I'm a researcher IRL and want empirical evidence: Show me any benchmark where any motherboard improves gaming performance significantly (5%+) in any game. Pro (MLC NAND) SSDs lasting longer is irrelevant, because even non-Pro SSDs (TLC NAND) last 100-400 years of ordinary use. Yes! Upgrade to 2.5 and you will get better performance. I have a 20% weaker graphics card than you, so if you're playing in 1920x1080 you should be able to max out all options in DCS 2.5 and still get minimum 40 FPS and average 50 FPS, unless you have an old CPU. 2.5 works great.

Hm, okay! Start the mission editor, create a new Caucasus scenario, put an A-10 right over the cross airfield south of Kobuleti, set skill to Player and start the scenario. After a few seconds when the camera is done zooming out, what framerates do you get? I looked at the FPS for about 20 seconds flying straight forward on autopilot and got 45-48 FPS.

I figured out what Full Screen is for. It didn't appear to do anything, so I recommended keeping it OFF, but what it does it that it keeps you in full screen even if playing in a low resolution and since some have reported Vsync problems, I'm going to change my recommendation to keeping it ON! I'll update the first post with this information. You mean you compared MSAA X2 and MSAA X4? Interesting! On my 1060, MSAA OFF to X2 and X2 to X4 cause -25% performance each, but X4 to X8 is practically free, which is weird and I don't know why this is, but I was expecting it to work differently on a weaker card, such as yours.

All specs, except CPU & GPU, are irrelevant to performance if they are above the minimum and with 20% CPU use and 100% GPU use on my machine, CPU is probably irrelevant, because the GPU is the bottleneck quite clearly. Do you have MSI Afterburner? I get 30 minimum and 30-35 FPS when I fly straight down the street with all the special buildings (south-to-north) in Vegas at 300 ft AGL and about 40 FPS flying around the same area, with 38 FPS being the number that’s is there the most when I watch the FPS. I’m playing the built-in “Take-off” mission, but I’ve opened the mission in the editor and put myself to start 4 NMI south of “Paradise”, or “Spring Valley”, or whatever the centre of Vegas is called, at 100 ft flying north. We don’t have to do this super scientifically, just see if there’s a difference of more than 5% and I’m also using 9GB RAM and 3GB VRAM (which is all I have). 1920x1200 has 10% more pixels than 1920x1080 and the performance difference is 10% in my benchmarks. I get about 45 FPS flying north along the road just west of McCarran and 33 FPS minimum as I pass straight through the centre, with the pyramid and stuff on my sides and almost flying directly into the Tromb building. My DCS 2.5.0 options are as follows: Textures HIGH Terrain Textures HIGH Civ. Traffic HIGH Water HIGH Visib. Range EXTREME Heat Blur HIGH Shadows HIGH Resolution 1920x1200 Aspect Ratio 16:10 Monitors 1 Res. of Cockpit Displays 1024 EVERY FRAME MSAA X8 Depth of Field OFF Lens Effects DIRT+FLARE HDR OFF Deferred Shading ON Clutter/Grass 1500 Trees Visibility 100% Preload Radius 100 Chimney Smoke Density 0 Gamma 2 Anisotropic Filtering X16 Terrain Object Shadows FLAT Global Cockpit Illumination ON Disable Aero Interface OFF Vsync OFF Full Screen OFF Scale GUI OFF My point is this though: the $800 worth of upgrades I describe above add ~0% performance. Motherboards don't do anything. i7's are i5's with hyper-threading, which most games aren't designed for and the difference is generally +/-5%, except choice games Water cooling isn't more effective than air cooling, it's cosmetic A RAM overhead literally does nothing, just like with unused HDD/SSD space Samsung EVO SSDs are just as fast as Samsung PRO SSDs if you're copying files smaller than 3-12GB (120GB-1TB version). Either way, it doesn't affect framerates or gaming performance AFAIK A PSU overhead may make it operate quieter, but any proposed advantages in energy efficiency and such are frivolous claims Spend the $800 on upgrading the graphics card from a 1060 to a 1070, or a 1070 to a 1080 instead and you'll get +20% performance, or a whopping 10 DCS modules! [pilotfly] And yes, there's also the element of cost-efficiency: each step in the 750, 760, 770, 780 ladder (from 2013) added like +20% performance, but the price doubled... currently, graphics card prices are all over the place due to cryptocurrency mining.

Hey, I have a 1060 3GB and was wondering if anyone’s got a 1060 6GB with an i5/i7 playing in 1080p/1200p, for comparison purposes? We could compare framerates flying over Vegas in the Warthog at the same options, or something. My theory is that you only need 3GB VRAM and that 6GB VRAM is a marketing ruse to get you to pay another $100, just like Gaming motherboards over ordinary motherboards (+$100), i7’s over i5’s (+$100), water cooling over air cooling (+$100), 16GB RAM over 8GB RAM (+$100), Pro SSDs (+$100), over-sized HDDs (+$100), over-powered supply units (+$100), and really, is there any honesty in the computer industry? The answer is no. The only honest computer component is the case. The thing with memory is that when a computer runs out of it, it usually makes do with what it already has and in theory, running out of it should probably only makes textures appear slower. I’ve never seen a benchmark demonstrate any significant (+5%) difference between cards with different amounts of VRAM, in 1080p anyway. May or may not wildly differ in 4K. I’ve compared a few 1060 3GB & 1060 6GB benchmarks and the FPS difference was small.