Jump to content

Pilotasso

Members
  • Posts

    11843
  • Joined

  • Last visited

  • Days Won

    10

Everything posted by Pilotasso

  1. It's coming this summer but it's the continuation of intel's policy to milk their customers for the same chip over and over again every year. You'll be paying a premium for an X299 motherboard and probably 400$ for a quad core chip that can be found on Z170 and Z270 platforms minus the crappy heat spreader. Thats right, likely the only difference.
  2. I tested my 1700X against sandy bridge, and kabylake in a thread next to this. The conclusion is that currently modern CPU's doesnt matter much as they once did, they all perform close enough for DCS. The GFX is likely what will differentiate systems apart for DCS performance. That Being said, you can always appreciate the extra threads Ryzen systems brings to the table for everything else beyond DCS so, that 1600X is my advice for you.
  3. EDGE has offloaded the geometry onto the GPU and since then the CPU has been under utilized except if you are getting a very high end GFX like a 1080Ti, where Sandy bridge might not have the grunt necessary to feed it anymore.
  4. this is very interesting. Could a GFX change bring both intel & AMD machines level despite the clock difference? wow, I think it can be done.
  5. Only FC3, A-10 and Mirage. Plan on getting the fast jets above anything else, WWII planes maybe in a bundle with discount.
  6. Images are worth a 1000 words. Whos gonna see this? looks stunning. :) gCcx85zbxz4
  7. Unfortunately when I ran DCS2 on my old 2500K I forgot to turn off MSAA and that introduced a GFX bottleneck. Even so I ran DCS2 again in these specific conditions with R7 overclocked and the results were: Mission start: R7=78 FPS Over nellis: R7=66 FPS Las vegas downtown:R7=37FPS Las vegas airport: R7=46FPS So comparing these results I made the chart below: Conclusion: Youll get at least 14% improvement going from sandy bridge to Ryzen with GFX bottle-necking in place and a 650Mhz disavantage!! Remember Im testing CPU+RAM differences here as the GFX is the same. But there's a caveat: looking at what others got with the 980Ti and 1080 I can safely assume that a new GFX like the 1080ti or VEGA would double or triple FPS, far exceeding what could be achieved with a processor change. In a final analysis, if you have sandy bridge and up and want to build for DCS specifically, just get a high end GFX. (I will verify this in 2 months) The only 2 variables capable of stopping this from being true is if you dont have enough RAM and no SSD. I can swear the later one is a BIG jump, just as important as a new GFX.
  8. @1440 with High preset VSYNC off with filters ANISO+MSAA off so that I could compare with the previous non overclocked Ryzen test which is quoted below. so I got 5-7 FPS just upping 450Mhz and switching from CAS 15 to 12.
  9. I overclocked the CPU to 3850Mhz, and tightened RAM timings to CAS 12-12-12-32-1T (kept @ 2933 because couldn't boot at 3200) results obtained: Mission start: 94 FPS Over nellis: 83 FPS Las vegas downtown: 47 FPS Las vegas airport: 53 FPS thats a 5 FPS increase across the board. But Las Vegas downtown upped by 7 FPS!
  10. strange, I didnt have to change anything and T.A.R.G.E.T. runs normally on my new R7 build.
  11. maybe something else was generating more threads?
  12. The last fighter program to be on schedule and budget was probably a WWII prop plane.
  13. well to minimize GPU bottleneck @1440 with High preset VSYNC off, I turned filters ANISO+MSAA off and then I get: Mission start: 89 FPS Over nellis: 77 FPS Las vegas downtown: 40 FPS Las vegas airport: 48 FPS Rage, maybe you can run these settings to compare as I cant go 4K :]
  14. Looks like those results reflect the GFX rather than the CPU. FPS wise you getting pretty nice scaling there: 970<980Ti<1080
  15. happy birth day!!! :yay: :drunk:
  16. OK, I ran task manager, terminated each executable at one time, and when I reached the culprit the problem went away. It was NZXT's CAM software FPS overlay. I only want to control the cooler's RGB but the software comes with alot of bloatware. removed cam software from the boot sequence, problem went away.
  17. Welcome to the hurdles of new PC building (see sig). Only this time the trouble is all on the operating system. I had to go thru uninstalling bloatware (Xbox DVR, telemetry, one drive, unwanted games/apps installed silently etc) But one that I cannot get rid off is this FPS counter that shows up on every game, tried everything (Gforce shadowplay GF experience, even trying older drivers) and nada. Some games wont run on full-screen like Mechwarrior online. Anyone knows how to get rid of the FPS counter??
  18. Yes, and yes. I'm waiting for BIOS updates in order to be able to overclock CPU & RAM more safely. People have reported many horror stories with current beta BIOSes, so Im keeping it very conservative. Graphics card will be a VEGA in a month or two. The GTX 970 is just a placeholder. by that time I expect the R7 pull away even if the increase will be rather modest.
  19. for those interested comparing results, see annexed track file: remember you must use same resolution and detail preset as I am. benchmark.trk
  20. Its benchmark time! OK guys so I ran DCS 2 over nevada. Test setup: Graphics settings used was the HIGH preset @1440P, with Vsync off so that doesnt get FPS capped. I5 System specs: 2500K@4.5Ghz 16GB RAM 1600MHZ CL 11 GTX 970 G1 Gaming stock Windows 7 R7 System specs: R7 1700X@3.5Ghz 16GB RAM @2933Mhz CL14 (I'm keeping the second kit in the box until BIOS updates) GTX 970 G1 Gaming stock windows 10 Here are the results: Mission start: 2500K=68 FPS, R7=71 FPS Over nellis: 2500K=58 FPS, R7=61 FPS Las vegas downtown: 2500K=30FPS, R7=33FPS Las vegas airport: 2500K=43 FPS; R7=41FPS The game reports very low GPU/CPU usage, TBH the results seem pretty much within margin of error. Kinda what I expected to be. Conclusions: ED better harness all this untapped hardware potential as the game optimization is falling behind what hardware can achieve these days. If your upgrading specifically for DCS and already have sandy bridge processors and up, save the money and get more memory, 8 or more GB and an SSD if you already dont have. Everything else is a waste of money. Below I post the screenshots for the timestamps above. The red FPS indication belongs to 2500K, the white FPS indication is R7 on windows 10 (for some reason it stays on on all games and cant get rid of it).
  21. No space on the desk as the pic below shows. I have it on the chair until I come up with a solution.
  22. lets see if these last photos came out decent with daylight.
  23. for DCS there is not much to gain from haswell to kaby lake. (for everything else: thats a completely different story) At this point I suspect DCS has been processor maxed out (due to engine, API and CPU tech limitations) and is rather increasingly GPU bound. In order to change this, ED would have to switch to DX12 and place stuff like a beefier flight model, missile, radar modeling and world events on separate cores and then crank the complexity of each thread up for added detail. But that wont happen any time soon.
×
×
  • Create New...