Jump to content

divinee

Members
  • Posts

    141
  • Joined

  • Last visited

Everything posted by divinee

  1. Didn't talk about the F-5E, only F-5. If i understood correctly the list was about aircraft which were in use at vietnam.
  2. You are missing F-4J and very early E. Edit: and F-5, F-8
  3. Hi, After the latest OB patch, there is still very heavy E-core usage and cpu's frametimes are in the 14ms region (in ST they are about 7-8ms). Also on P-cores, there seem to be load only on one thread per core. Tested in the hornet's ready on the ramp caucasus as in the OP. dcs.log DxDiag.txt
  4. I might be in love with that one from VF-96..
  5. I think E-phantom has ASQ-19 modified version of ARC-51, RT-793/ ASQ
  6. Upgraded Intel Management Engine firmware to the version which came out at january and the core usage looks much better now. There is still quite a lot e-core usage but now there is 15-45% load on every P-core. Weird thing is that mostly only one thread per core is still used even though i have hyperthreading enabled. Also the frametimes didn’t go down where they suppose to be. Sorry for the quality of the pic..
  7. Made some more testing today. Any affinity tricks doesn’t work for me. Then just for humor i tried to change windows power plans from control panel on the fly. When changing between best performance and balanced there were zero change. After switching to power saver, the cpu frametimes went down to 3-10ms. I have to do some more testing but someone with this problem can propably try that also to confirm my findings? Edit: before testing, just make sure that cpu doesn’t drop to low power mode by forcing the clocks.
  8. Hi, Thank you for the reply. Followed instructions provided in that topic and information attached. dcs.log DxDiag.txt
  9. I would even pay more if i'd have to stress about the reliability of the systems and if my missiles, radar etc just randomly wouldn't work . It's a big part of the characteristics of those machines and I'm flying DCS because of the realistic simulation, not to be the most efficient target killer. Of course not everyone wan't those so those should be optional things (from the server side).
  10. 75-80 degrees in single threaded is very hot for that cpu. Should be around 60-70 in single threaded to handle multithreaded even reasonably.
  11. Please, put a note in the post if you are running with 2d or vr so we can rule out possible causes.
  12. I also tried to block E-cores with Lasso and the load clearly moved to P-cores. I didn’t notice any change in the frametimes though..
  13. Seems like i found a solution. There was high performance selected in windows power plan. Changed it to balanced and cpu-frametimes dropped to 3-10ms in the same scenario as in the original post. Also there seems to be more even load to P-cores. edit: nah, might be a false alarm. Restarted everything and frametimes went back to 16-17ms. I think i somehow managed to trigger windows core scheduling to notice that dcs is gaming workload, not utility load. But still changing the power plan was a step to the right direction. @BIGNEWY
  14. Thank you for the reply. I did all those in said order, but saw no change in the situation. New log file attached. edit: and DCS still refuses to start if i disable all e-cores from bios. I cannot find any info anywhere what can cause this kind of behavior. dcs.log
  15. This is very interesting. Will be following this because i have similar problem with Intel p/e-core processor which is using only e-cores for DCS MT. These can be possibly connected. https://forum.dcs.world/topic/321019-mt-running-with-e-cores-only/
  16. I think i’m having the same problem. I made a bug report in MT area if you want to check?
  17. Hi, Seems that DCS is running with E-cores on my system. This causes massive increase in frametimes. With ST CPU frametimes in hornet's ready on the ramp instant action were about 8ms. With MT CPU frametimes increased to ~16-17ms. After parking all the e-cores, dcs refuses to start at all. System is on my sig. Forgot to mention that running with native OXR and G2 with 100%. edit: tried slow repair but it didn’t help edit2: noticed that when dcs starts it loads up one P-core. Same thing when loading mission. After load is done all the P-cores are idling and E-cores are doing all the work. DxDiag.txt dcs.log
  18. Nope and nope. J is newly built version with big upgrades to structure and systems including awg-10. J and B are different generations of the naval phantoms. N is rebuilt B with small upgrades derived from J but it still has the apg-72 from the B. S is overhauled and improved J to handle the biggest problems like the questionable reliability of older, more analog awg-10, maneuverability, smokeless engines (some J’s already had these), more radios. Good source for the quick info which seems to be more correct than wikipedia: http://www.joebaugher.com/usaf_fighters/f4.html
  19. I’m happy with the last update. If i send feedback to ED it doesn’t mean i’m not happy with the update. Mostly people who doesn’t have any problems don’t post that much here. There is few quirks but they are solvable and ED seems to be on top of it.
  20. Turbo mode caused freezing for me. After switching off there haven’t been any problems.
  21. I'm very happy with all the variants we get altough i'd favor naval variants. Will be lots of fun with all those
  22. No, it's F-4 forum which includes all the models Heatblur will make including naval models.
×
×
  • Create New...