War Thunder — Games you keep reinstalling every year

Page 1 of 2 · 24 posts

1. ISIris S.

   Iris S.Original post

   @iris_s · 1 Jun 2026, 00:17

   #1

   Not a complaint, just genuinely puzzled by the behaviour. The default settings are clearly tuned for a different setup than mine.

   Half the advice online is a year out of date, which is why I wanted a current thread. Will update the post if I find a cleaner solution.

2. ESElla S.

   Ella S.

   @ella_s · 3 Jun 2026, 13:39

   #2

   Worth checking first: that behaviour is intentional, even though it is poorly explained in-game.

3. MEMercurypulse

   Mercurypulse

   @mercurypulse42 · 5 Jun 2026, 12:11

   #3

   Only partly related, but temperature is the usual culprit when performance degrades over a session.

4. ASAstralnyx

   Astralnyx

   @astralnyx99 · 6 Jun 2026, 16:34

   #4

   From experience, the safe move is the reversible one — you can always undo it if it makes things worse.

5. MPMika P.

   Mika P.

   @mika_p10 · 7 Jun 2026, 16:14

   #5

   Slightly different take — do the cheap diagnostic first — it costs nothing and rules out half the causes.

   Worth a shot before anything drastic.

6. SASaberflare

   Saberflare

   @saberflare85 · 10 Jun 2026, 22:31

   #6

   From experience, measure before and after, otherwise you are just guessing whether it helped.

7. BLBlazenyx

   Blazenyx

   @blazenyx15 · 11 Jun 2026, 06:23

   #7

   For what it is worth, if it only happens in one mode, that narrows it down considerably.

8. DEDeltamuse

   Deltamuse

   @deltamuse · 13 Jun 2026, 13:43

   #8

   I went through this last month. nine times out of ten it is a driver or a background process, not the thing you suspect.

9. ARArcbyte

   Arcbyte

   @arcbyte94 · 15 Jun 2026, 14:03

   #9

   Quick correction to the above: temperature is the usual culprit when performance degrades over a session.

10. PGPriya G.

    Priya G.

    @priya_g · 17 Jun 2026, 05:32

    #10

    The usual fix here is you get most of the benefit from the first adjustment; everything after that is marginal.

11. EKElla K.

    Ella K.

    @ella_k15 · 17 Jun 2026, 09:51

    #11

    Depends on your setup, but broadly the trick is to change one thing at a time and actually note the result before touching anything else.

12. EZElla Z.

    Ella Z.

    @ella_z6 · 19 Jun 2026, 08:37

    #12

    Speaking as someone who got it wrong initially, the averages hide the problem; look at the one percent lows instead.

13. NONovawolf

    Novawolf

    @novawolf17 · 22 Jun 2026, 13:33

    #13

    This comes up a lot, so: the defaults are conservative on purpose, so there is usually headroom to gain.

14. PHPhantomvale

    Phantomvale

    @phantomvale70 · 22 Jun 2026, 23:39

    #14

    Only partly related, but the community guide covers this better than I can, but the summary is to prioritise the basics first.

    Others may have a cleaner method.

15. ADAya D.

    Aya D.

    @aya_d · 26 Jun 2026, 09:17

    #15

    Speaking as someone who got it wrong initially, the trick is to change one thing at a time and actually note the result before touching anything else.

16. CICinderlynx

    Cinderlynx

    @cinderlynx81 · 26 Jun 2026, 15:00

    #16

    This is the bit people miss — give it a clean reinstall before assuming the hardware is at fault.

17. LNLeo N.

    Leo N.

    @leo_n · 27 Jun 2026, 22:44

    #17

    Adding to the above: it is almost always a setting rather than hardware, so try the free fixes before spending anything.

18. NENoa E.

    Noa E.

    @noa_e · 1 Jul 2026, 02:26

    #18

    Counterpoint: it is almost always a setting rather than hardware, so try the free fixes before spending anything.

19. DBDeniz B.

    Deniz B.

    @deniz_b29 · 3 Jul 2026, 12:45

    #19

    Short version: you get most of the benefit from the first adjustment; everything after that is marginal.

20. CRCrimsonash

    Crimsonash

    @crimsonash · 5 Jul 2026, 06:23

    #20

    Not the answer you want, but region and server load matter more than most guides acknowledge.