159th_Falcon Posted June 21, 2011 Posted June 21, 2011 (edited) Ok, got a feeling something doesn't add up here. My system; Phenom II 965 BE ASUS M3A32MVP Deluxe Kingston HyperX RAM (8 GB) Radeon 9650 2GB All on W7 64 Now, whenever my AC is on the ground my fps won't go above 20'ish. Then, when i take a look at CPU/GPU usage neither is maxed out. (looking at a single core for the CPU ofcourse) What's going on? Shouldn't there always be either a GPU or CPU core be maxed out? Once i am in the air i usually have 30 fps which is where i set the limit. Though, on occasion it drops down to 25 orso, whitout a change in GPU/CPU usage? Anyone has an idea why this is and what's going on? And how to change it if possible? ~S~ ill see if i can provide a screenshot late in the evening. Showing both CPU/GPU usage and ingame FPS. ***EDIT*** Talking A10C btw Edited June 21, 2011 by 159th_Falcon specified game A10C [sIGPIC][/sIGPIC] The keeper of all mathematical knowledge and the oracle of flight modeling.:)
EtherealN Posted June 21, 2011 Posted June 21, 2011 A DxDiag would be interesting. I suspect your GPU has slow memory bus, preventing it from using it's full compute potential, but I haven't been able to find any data on a Radeon 9650. Are you sure that's what you have? [sIGPIC][/sIGPIC] Daniel "EtherealN" Agorander | Даниэль "эфирныйн" Агорандер Intel i7 2600K @ 4.4GHz, ASUS Sabertooth P67, 8GB Corsair Vengeance @ 1600MHz, ASUS GTX 560Ti DirectCU II 1GB, Samsung 830series 512GB SSD, Corsair AX850w, two BENQ screens and TM HOTAS Warthog DCS: A-10C Warthog FAQ | DCS: P-51D FAQ | Remember to read the Forum Rules | | | Life of a Game Tester
Drona Posted June 21, 2011 Posted June 21, 2011 I think he means the HD Radeon 6950 because the Radeon 9650 is an AGP card for the PowerMac G5 :).
EtherealN Posted June 21, 2011 Posted June 21, 2011 Yeah, but I'd need that confirmed if possible. :) [sIGPIC][/sIGPIC] Daniel "EtherealN" Agorander | Даниэль "эфирныйн" Агорандер Intel i7 2600K @ 4.4GHz, ASUS Sabertooth P67, 8GB Corsair Vengeance @ 1600MHz, ASUS GTX 560Ti DirectCU II 1GB, Samsung 830series 512GB SSD, Corsair AX850w, two BENQ screens and TM HOTAS Warthog DCS: A-10C Warthog FAQ | DCS: P-51D FAQ | Remember to read the Forum Rules | | | Life of a Game Tester
Drona Posted June 21, 2011 Posted June 21, 2011 Yeah, but I'd need that confirmed if possible. :) :doh: Understood.
159th_Falcon Posted June 21, 2011 Author Posted June 21, 2011 (edited) :doh::doh: 6950 yes, for some reason i always mix up the 6 and the 9 A Saphire 6950 2GB Flex edition using the latest drivers provided by Saphire. Even though the 69 is such a lovely.... Oh wait, different subject. Anyway, how to get that DxDiag report? Edited June 21, 2011 by 159th_Falcon GFX card brand is Saphire, not Asus as posted previously [sIGPIC][/sIGPIC] The keeper of all mathematical knowledge and the oracle of flight modeling.:)
EtherealN Posted June 21, 2011 Posted June 21, 2011 Ah, if it's a 6950 then I don't think a memory bus speed issue is the likely cause. To get a DxDiag, press Start, then type "dxdiag" in the search field and start that application. THen use the "save to file" feature to get a dxdiag.txt file. Attach that here through the "go Advanced" button and the paperclip icon. [sIGPIC][/sIGPIC] Daniel "EtherealN" Agorander | Даниэль "эфирныйн" Агорандер Intel i7 2600K @ 4.4GHz, ASUS Sabertooth P67, 8GB Corsair Vengeance @ 1600MHz, ASUS GTX 560Ti DirectCU II 1GB, Samsung 830series 512GB SSD, Corsair AX850w, two BENQ screens and TM HOTAS Warthog DCS: A-10C Warthog FAQ | DCS: P-51D FAQ | Remember to read the Forum Rules | | | Life of a Game Tester
159th_Falcon Posted June 21, 2011 Author Posted June 21, 2011 Ok, here they are. I saw there was an option to run an 64bit DxDiag so attacheched that one as well, no idea if it matters though. thanx for the help so far.DxDiag.txtDxDiagx64.txt [sIGPIC][/sIGPIC] The keeper of all mathematical knowledge and the oracle of flight modeling.:)
Drona Posted June 21, 2011 Posted June 21, 2011 159th Falcon, I'm just curious as to what simulation you are flying? A-10C or FC2?:joystick:.
159th_Falcon Posted June 21, 2011 Author Posted June 21, 2011 (edited) Both along whit the shark, but topic applies to A10C. ***EDIT*** Just took some screenshots where you can see GPU/CPU/RAM usage. Something i just noticed is that the FPS doesn't change when i pause the game?? 1st and 2nd picture taken whit game paused, 3rd and 4rd picture taken whit game unpaused. Edited June 21, 2011 by 159th_Falcon Added Screenshots [sIGPIC][/sIGPIC] The keeper of all mathematical knowledge and the oracle of flight modeling.:)
EtherealN Posted June 21, 2011 Posted June 21, 2011 Just took some screenshots where you can see GPU/CPU/RAM usage. Hm, a more detailed look at the GPU load would be necessary to say anything definitive. That's a block diagram of a Cayman GPU like the 6950. Depending on what's going on and how the GPU load is measured by your program, it is entirely possible to have the GPU tasked to capacity but still show less utilization than 100%. For example, if 100% assumes that all components are fully tasked, but the simulator doesn't use everything (The Tesselator for example), it might register as less than 100%. Hard to say for sure though - would need someone with more AMD/ATi experience to say for sure. Benchmarks indicate, however, that you should be getting much better performance than you are experiencing even at max settings if it is the GPU that is causing the issue - it's roughly equivalent to my 560Ti and it does a lot better than that with max graphics in windowed mode, Aero active and two screens running... The Phenom 2 should perform better as well, in my opinion. Something i just noticed is that the FPS doesn't change when i pause the game?? Shouldn't necessarily happen anything with the FPS just because you pause the game. You are essentially pausing time, but you are not pausing rendering or any of that. I don't know if the simulation engine itself pauses, but I suspect it doesn't. I'd recommend that you create a track file where you perform a set of actions - say, taxi around a bit with other planes in the vicinity or whatever. Then play those while recording FPS with the FRAPS benchmarking feature and test it out with different settings. Exit and close the simulator between each test run. (To avoid issues with caching polluting your data.) Perhaps you can find something that's "doing it". We probably need to find you someone that is more familiar with AMD hardware though - they might know some tweak that might cause trouble without being obvious. [sIGPIC][/sIGPIC] Daniel "EtherealN" Agorander | Даниэль "эфирныйн" Агорандер Intel i7 2600K @ 4.4GHz, ASUS Sabertooth P67, 8GB Corsair Vengeance @ 1600MHz, ASUS GTX 560Ti DirectCU II 1GB, Samsung 830series 512GB SSD, Corsair AX850w, two BENQ screens and TM HOTAS Warthog DCS: A-10C Warthog FAQ | DCS: P-51D FAQ | Remember to read the Forum Rules | | | Life of a Game Tester
159th_Falcon Posted June 21, 2011 Author Posted June 21, 2011 Alright, thanx for the input EtherealN. For the record, im running 2 24" monitors @ 1920*1200 resolution. One for the game and one i keep as desktop, as you can see in previous screenshots. In A10C i have pretty much every setting on max; TEXTURES: HIGH SCENES: HIGH CIV TRAFFIC: HIGH WATER: HIGH VISIB RANGE: HIGH HEAT BLUR: ON SHADOWS: HIGH RESOLUTION: 1920*1200 ASPECT RATIO: 16:10 MONITORS: 1 SCREEN RES. OF COCKPIT DISPLAYS: 1024 MSAA: 8xQ HDR: HARD CLUTTER/BUSHES: 500M TREES VISIBILITY 7026M VSYNC ON FULL SCREEN OFF COCKPIT SHADOWS ON TSAA ON G-EFFECTS SIMULATION MIRRORS ON Thing is though, fps does not seem to be affected regardles of the GFX settings i use. So im guessing its something to do whit the CPU?? Anyway, someone knows a program so we can get a more accurate picture of the GPU load? [sIGPIC][/sIGPIC] The keeper of all mathematical knowledge and the oracle of flight modeling.:)
EtherealN Posted June 21, 2011 Posted June 21, 2011 Could you grab load pictures of the same situation, but with everything set to either off or as low as it gets? Plase disable the FPS limiter as well. [sIGPIC][/sIGPIC] Daniel "EtherealN" Agorander | Даниэль "эфирныйн" Агорандер Intel i7 2600K @ 4.4GHz, ASUS Sabertooth P67, 8GB Corsair Vengeance @ 1600MHz, ASUS GTX 560Ti DirectCU II 1GB, Samsung 830series 512GB SSD, Corsair AX850w, two BENQ screens and TM HOTAS Warthog DCS: A-10C Warthog FAQ | DCS: P-51D FAQ | Remember to read the Forum Rules | | | Life of a Game Tester
Kuky Posted June 22, 2011 Posted June 22, 2011 (edited) hm... why are you running windowed mode if you are running 1 screen only? I think you are pushing the graphics side a bit too much, I mean you have 8xAA, TSAA on, HDR on, Mirros on, Grass to 500M etc etc and High vis distance... that's bit too much to have very good FPS in my opinion. I think just reducing AA to 4x instead of 8x and reducing Grass to 300-350m or less (or less if you are using my dense grass tweak) and HDR off and you'll almost double the FPS I think. And Ethereal is right, that 100% utilisation of the GPU should definitely be showing only when certain part of the GPU core, it won't show if shaders or part that does TSAA or memory buss is maxed out so my bet is one of those things is maxed out before main core of the GPU is maxed out. One more tip though... turn TSAA off in game and use Adoptive AA in ATI driver, this should also give some FPS boost, and don't use the highest quality TSAA in graphics driver, that really impacts FPS also. Edited June 22, 2011 by Kuky PC specs: Windows 11 Home | Asus TUF Gaming B850-Plus WiFi | AMD Ryzen 7 9800X3D + LC 360 AIO | MSI RTX 5090 LC 360 AIO | 55" Samsung Odyssey Gen 2 | 64GB PC5-48000 DDR5 | 1TB M2 SSD for OS | 2TB M2 SSD for DCS | NZXT C1000 Gold ATX 3.1 1000W | TM Cougar Throttle, Floor Mounted MongoosT-50 Grip on TM Cougar board, MFG Crosswind, Track IR
EtherealN Posted June 22, 2011 Posted June 22, 2011 hm... why are you running windowed mode if you are running 1 screen only? He's running two screens. ;) [sIGPIC][/sIGPIC] Daniel "EtherealN" Agorander | Даниэль "эфирныйн" Агорандер Intel i7 2600K @ 4.4GHz, ASUS Sabertooth P67, 8GB Corsair Vengeance @ 1600MHz, ASUS GTX 560Ti DirectCU II 1GB, Samsung 830series 512GB SSD, Corsair AX850w, two BENQ screens and TM HOTAS Warthog DCS: A-10C Warthog FAQ | DCS: P-51D FAQ | Remember to read the Forum Rules | | | Life of a Game Tester
Kuky Posted June 22, 2011 Posted June 22, 2011 He's running two screens. ;) hm... I understod he's got 2 screens connected but runs 1 screen for the sim only... well if he's using 2x 1920x1200 then he's definitely maxing out his video card, I have no doubt about that PC specs: Windows 11 Home | Asus TUF Gaming B850-Plus WiFi | AMD Ryzen 7 9800X3D + LC 360 AIO | MSI RTX 5090 LC 360 AIO | 55" Samsung Odyssey Gen 2 | 64GB PC5-48000 DDR5 | 1TB M2 SSD for OS | 2TB M2 SSD for DCS | NZXT C1000 Gold ATX 3.1 1000W | TM Cougar Throttle, Floor Mounted MongoosT-50 Grip on TM Cougar board, MFG Crosswind, Track IR
EtherealN Posted June 22, 2011 Posted June 22, 2011 hm... I understod he's got 2 screens connected but runs 1 screen for the sim only... well if he's using 2x 1920x1200 then he's definitely maxing out his video card, I have no doubt about that See the screenshots he posted - he's essentially running it the same way I am (though I have my secondary screen on the left instead of the right). But yes, it would be interesting to compare performance with only one screen operational (quickly achieved through Win-P). [sIGPIC][/sIGPIC] Daniel "EtherealN" Agorander | Даниэль "эфирныйн" Агорандер Intel i7 2600K @ 4.4GHz, ASUS Sabertooth P67, 8GB Corsair Vengeance @ 1600MHz, ASUS GTX 560Ti DirectCU II 1GB, Samsung 830series 512GB SSD, Corsair AX850w, two BENQ screens and TM HOTAS Warthog DCS: A-10C Warthog FAQ | DCS: P-51D FAQ | Remember to read the Forum Rules | | | Life of a Game Tester
Udat Posted June 22, 2011 Posted June 22, 2011 Difference between ground and air; shouldn't that be the grass rendering? I would try turning grass rendering off completely and check if the difference is still there. Intel i7-950 @stock, Asus P6X58D-E, 3x4GB Corsair Vengeance, Asus GTX 580, Corsair 120GB SSD, Corsair HX 750W PSU [sIGPIC][/sIGPIC]
Kuky Posted June 22, 2011 Posted June 22, 2011 Both along whit the shark, but topic applies to A10C. ***EDIT*** Just took some screenshots where you can see GPU/CPU/RAM usage. Something i just noticed is that the FPS doesn't change when i pause the game?? 1st and 2nd picture taken whit game paused, 3rd and 4rd picture taken whit game unpaused. I had a look at those screenshots, so he's still running 1 screen only with 1 monitor (second monitor is showing desktop only) so A-10C is running on single monitor... he needs to turn full screen and not run windowed mode. One thing to do now is while sim is running go into windows task manager and set affinity for DCS.exe to single core only then watch that utilisation, I am certain it will be 100% or close to that and sim will run at even les FPS (because sound is running off the same core), if you switch affinity to 2 or more cores it'll run same FPS as before but then you can't see exactly how much of single core DCS is using because it's load is spread out. PC specs: Windows 11 Home | Asus TUF Gaming B850-Plus WiFi | AMD Ryzen 7 9800X3D + LC 360 AIO | MSI RTX 5090 LC 360 AIO | 55" Samsung Odyssey Gen 2 | 64GB PC5-48000 DDR5 | 1TB M2 SSD for OS | 2TB M2 SSD for DCS | NZXT C1000 Gold ATX 3.1 1000W | TM Cougar Throttle, Floor Mounted MongoosT-50 Grip on TM Cougar board, MFG Crosswind, Track IR
Recommended Posts