Jump to content

Groove_C

Member
  • Content Count

    940
  • Joined

  • Last visited

  • Medals

Everything posted by Groove_C

  1. @Tankbuster @ 1440p (3.7 Mpix) vs. 2160p (8.3 Mpix) ultra preset GTX 1080 delivers ~2x more FPS in certain areas/scenarios. So 150% 3D resolution is not a good idea. @ 1440p ultra preset RTX 2070/2080 vs. GTX 1080 delivers only 3-4 FPS avg. more, in singleplayer. If I were you, I wouldn't replace the GTX 1080. @ 1080p standard preset 12k view distance DDR4 2666 MHz i9-9900K (4.7 GHz 8C/16T) vs. i7-9700K (4.6 GHz 8C) delivers 5 FPS avg. more, in singleplayer. @ 1080p standard preset 12k view distance i7-4790K (4.7 GHz @ DDR3 1600 MHz) is equal to i7-7700K (4.4 GHz @ DDR4 2666 MHz) @ 1080p standard preset 12k view distance DDR4 2666 i7-9700K (4.6 GHz 8C) vs. i7-7700K (4.4 GHz 4C/8T) delivers only 5 FPS avg. more, in singleplayer. @ 1080p standard preset 12k view distance i7-9700K (4.6 GHz @ DDR4 2666 MHz) vs. i7-4790K (4.7 GHz @ DDR3 1600 MHz) could add only 5 FPS avg. more, in singleplayer. @ 1080p standard preset 12k view distance i7-9700K (5.0 GHz @ DDR4 3600 MHz) vs. i7-4790K (4.7 GHz @ DDR3 1600 MHz) could add whopping 17 FPS avg. more, in singleplayer. But if you play @ 1440p very high/ultra preset with <12k view distance, avg. singleplayer FPS gain would be < or = 10. In multiplayer the difference would be even smaller. My suggestion would be to lean towards the i9-9900K. Look now at people that were thinking back then, that a 4 cores i5 is more than enough (@ very high frequency). The most of them have already (had to) switched to newer platforms (vs. i7 owners) or have kept them and are dealing with less FPS and/or more micro stuttering in games. The i9-9900K, because of its 16 threads is more future proof. More and more games are starting to use 8 threads and more and I think the years to come, their number will only increase. Regarding ArmA only, I have seen some "finally correct" CPUs' comparisons on german forums @ lower frequencies but with more cores, where not only the min./max./avg. FPS was compared, but the frame time as well! The thing is, that if a game can't properly use/load a CPU with 12/16 threads, it doesn't mean that there is no benefit of having more threads. The most of well known sites continue to use only FPS as comparison, but it's outdated and doesn't tell the whole story! Frame time is the key! FPS is the speed of the game, but frame time stability and how low it is, is the smoothness of the game. So one can have 80 FPS with 4 cores i3-8350K @ 4.8-5.0 GHz and despite the FPS, there will be micro stuttering. And one can have 50 FPS with 12/16 threads Ryzen @ 4.0-4.2 GHz and despite the FPS, everything will feel butter-smooth. And this is because of a lot of background tasks being executed, that interfere with the game by accessing/loading the same CPU threads that the game does. The more threads your CPU has, across more of them can the background tasks be split and this way there is less probability, that certain background tasks will access particular threads, being used by the game or if, then there will be less load by these background tasks. In short, you won't get that much more FPS at your resolution/quality preset, in multiplayer (certainly not for $$$ + considering your curent setup), but it will be much more stable and the most important a lot smoother! But I wouldn't buy the i9-9900K now. A "newer" i9-9900K version is already available - the i9-9900KF (disabled iGPU), result of the same, but "apparently" slightly improved 14 nm production, with "apparently" slightly better IMC (integrated memory controller) and "apparently" overclocks ~100 MHz better or same frequency as the i9-9900K, but requires slightly less voltage. Ryzen 2 unveiling in May. Maybe the regular i9-9900K will be discounted after Ryzen 2 release. Coffee Lake re-refresh CPUs in Q2. A 650 Watt gold modular Corsair RM650X PSU is more than enough for a heavily overclocked i9-9900K and RTX 2080 Ti . A motherboard with enough/cool VRMs to feed an overclocked i9-9900K is required, like Gigabyte Aorus Master. 3600 MHz CL15 G.Skill Trident Z would be a very good choice. 3600 MHz CL17 G.Skill Trident Z is even "better" (if you know well how to overclock by lowering/upping latencies/frequencies and increasing VDIMM, SA and IO voltages).
  2. @Tankbuster i7-4790K frequency? RAM frequency, amount and latencies? Do you record your gameplay? Which software for recording? CPU or GPU recording? Do you edit videos? Do you know how to overclock CPUs? Do you plan to overclock your upcoming CPU?
  3. @Mahatma Gandhi I have finally finished optimizing secondary and tertiary timings of my 32 GB Trident X RAM. And here is the result @ standard preset @ 1080p @ 4.8 GHz (instead of 5.0 GHz).
  4. Groove_C

    1.90 update - Troubleshooting

    Dozens of other servers (with and without mods) on the other hand, simply work and have more players than usually, that have discovered them, because of their favorite servers being down. And as a lot of servers run similar missions/mods, some of players will probably stay on other servers even after the fix and won't come back to the servers where they've always plaid before. Very frustrating.
  5. Groove_C

    1.90 update - Troubleshooting

    @RH-SPECTER The thing with never ending "loading data" we have several times a week on our server after server restart. When players connect, they can be in lobby just fine, but when clicking OK, it says "loading" and nothing happens. And you cant wait for eternity with no results. Only 1 or 2 server restarts "solve" this. Started to happen after 1.88.
  6. Guys, please check this and tell me if some of you have also this problem or know what this could be or how to solve it.
  7. I have tested client RAM consumption 1.90 vs. 1.88. After 10 hrs of uninterrupted MP gameplay (no reconnect/restart + server restart for this purpose) I can say that now client RAM consumption is much lower. 16-17 GB vs. 22-23 GB + better server FPS for the same duration.
  8. Groove_C

    1.90 update - Troubleshooting

    There is also a problem that appeared exactly with 1.88 (nothing changed in the mission/server + no mods) that we have occasionally on our server and have reports from 2 other servers, running wasteland. When any player gets in a random (already spawned by the mission) car/helicopter - it simply explodes. If you don't get in - it doesn't explode. It doesn't concern vehicle(s) you purchase, but only certain vehicle(s) that is/are spawned by the mission. After the mission launch everything can be fine or there can be this one vehicle or more than one, that will explode each time somebody gets in, after the vehicle respawns back at the base. And at each new mission start one never knows if there will be this problem or not and if yes, for which of the vehicles present at the base. When these particular cars/helicopters explode, it will do this only for them, till the mission resatrt, no matter who gets in and when. I have deleted some of them with "deletevehicle cursortarget" and have driven at the place where they respawn. And I have witnessed and INVISIBLE OBSTRUCTION on the ground that you can collide with. So this invisible obstruction can appear or not, for only one vehicle or for more than one. Never seen this before and other server owners don't know how to solve this as well, as nothing was changed in our different missions, running on different servers. We haven't added anything invisible in our missions to make vehicle(s) explode on purpose by colliding/interfering with them to sabotage game of our players. Definitely nothing to do with our different missions/servers as everything has always worked just fine for a long time. And one never touches/changes things that have proven to work well. Somebody knows what could be the problem here?
  9. Groove_C

    1.90 update - Troubleshooting

    Have also check our RPT log and can also see: BEServer::finishDestroyPlayer(225296211): users.get failed BEServer: cannot find channel #225296211, users.card=19 But players still can join the server, just not that much players on our server as on saturday. No mods + 2 HCs. But starting from yesterday we have problems reviving some players as it's not always working. They are revived, but stay unconscious no matter how many times they get revived by the same player or other players. Nothing was changed in the mission or server. + spam in the chat coming from these problematic players in this form:
  10. +1 I have purchased mine and was ready to return it in case if it wouldn't have booted, as it's not listed in the QVL for my mobo. Good to know:
  11. Unlisted RAM only means that this particular RAM was not tested for compatibility, but not that it's not compatible. Consider, that new RAM comes all the time, even after the mainboard is not produced/updated anymore. It's simply almost impossible to test all available RAM on all available mainboards. My 32 GB (4X8) 2400 MHz G.Skill RAM (1.65 V 10-12-12-31-2) is also "not supported" by my mainboard. But I have successfully tested it with 8 hrs Prime95 and 8 hrs Karhu RAM Test at <1.65 V and 10-11-12-18-1 primary timings. No crashes in ArmA even after >8 hrs of non-stop gameplay. Secondary timings are also set. Have just to finish configuring/testing remaining tertiary timings (marked in red). (ASRock Timing Configurator)
  12. @MuffinDealer 1. 3200 MHz is the absolute max RAM frequency for 1st gen. Ryzen top high end mainboards. 2. 1st gen. Ryzens are not that high RAM frequencies friendly (only a minority of them are capable of 3200 MHz). 3. You should buy Ryzen compatible RAM (should be single rank?). 4. Your RAM should be placed in A1 and B1 slots (1 slot of space between them for dual channel operation) 5. You should update your BIOS to the latest available version (December 2018), which includes AGESA (AMD microcode update) 1.0.0.6 version for improved RAM stability/compatibility. 6. You should set you RAM to XMP profil in the BIOS. 7. There is no problem @ 3200 Mhz with 2nd gen. Ryzens with the latest AGESA using Ryzen compatible RAM. Othewise try to find the lowest latencies for 2933 MHz. 2933 MHz dual rank RAM is on par with single rank 3200 MHz RAM. Yours is dual rank, I think.
  13. And eDRAM compensates a lot (FPS) for not having the fastest RAM available, because with DDR3 1600 CL9 vs. DDR3 2400 CL9, there FPS difference can only be measured, but not noticed at all, using an i7-5775C.
  14. i7-5775C still delivers a very good performance in CPU limited games. 4.2/3.3/1.8 GHz Core/Cache/eDRAM. 1600 MHz DDR3 CL9. Cache/eDRAM could have been overclocked to 3.6-3.8/2.0-2.2 GHz. So imagine 16 threads @ 5.0+ GHz with eDRAM )))
  15. Groove_C

    [MP][CTI-COOP] EvorA

    @Sgt. Dennenboom I have contacted somebody who knows the server owner and he apparently has already removed your mission from his server, where the original mission name was changed and all authors removed/replaced. Your mission is being used on this server: In my understanding, russian translation is done using google translate, therefore not accurate and not everything was translated. So I wouldn't recommend you to include/take this translation 🙂 It would have been much better if you would have "moved" earplugs and weapon holstering to the action menu. Also what does EvorA actually mean? Evo red Altis? It's really not a good name for the mission, because one doesn't understand what kind of mission it is judging mainly by its name. Very often people search using "evolution" in the launcher/in-game browser or in battlemetrics and this way the mission can't be found. This is one of the reasons why almost nobody plays the mission. You could use instead a "better" naming for the mission for "better" recognizability, like Evolution CSAT vs. NATO + removing the name of the island as it's mentioned in the browser anyways.
  16. Groove_C

    [MP][CTI-COOP] EvorA

    @Sgt. Dennenboom Today I have found a server that runs your mission with mission name completely changed and authors removed everywhere and replaced as if the mission was done by somebody else. Mission weights ~2.7 MB, with a lot of script errors. I have the .pbo of the modified mission if you need it.
  17. I have discovered a few "game breaking" bugs. - If you get in as commander in a vehicle, but you're not the first to get in the vehicle - there are no smokes available to use. You have to be the first to get in the vehicle and as commander to have smokes available. - In Kayman, when you order the AI to switch to HE cannon ammo, it switches to APDS and vice versa.
  18. Groove_C

    [MP][CTI-COOP] EvorA

    @Sgt. Dennenboom
  19. Groove_C

    [MP][CTI-COOP] EvorA

    When to expect the update +-?
  20. Groove_C

    [MP][CTI-COOP] EvorA

    @Sgt. Dennenboom Is it ok if the mission will be translated to russian and changes will be made to CSAT/NATO equipment/vehicles? You should seriously consider to add earplugs. Vehicles one purches should be locked by default. Lock vehicle button shouldn't push other players out, as there can be players that are wanted as crew. Instead add a separate "push crew out" button.
  21. @Mahatma Gandhi It's a guy from overclock.net. There are even a few, that have their i7-5775c @ 4.4/4.0/2.4 GHz core/cache/eDRAM :-)
  22. @Mahatma Gandhi An other i7-5775c 4.2/3.8/2.0 GHz core/cache/eDRAM + 2x8 GB 2400 MHz dual rank 11-12-12-32 RAM ;-)
  23. @3D_vet Lets first see how much they will cost. I'm afraid that it could become more expensive if IPC and frequency gains will be true to what's being discussed currently.
×