How many of you are using multi monitor setups?
-
Yes Arty I use them already but still the beta version because the latest didn’t support my secondary card anymore. I use a 5450 for my MFDs. I think it runs better with the new drivers.
But the hunt for “TGP frames” continues. Have to test again what happens without shadows, shaders etc. -
Buhhhh although eyecandy I consider them a must cause in real life visuals are way more plural than the sim. So those eyecandy are a must for enjoyment and I refuse to let them go. I prefer to spend some more money and have them than forget about them to have more fps.
Are you OCing the beast?
IMHO your fps are great. Now I’m on the 20ish, rarely viewing 30+ and my experience is very fluid.
Today I get a new (used) hw combo just to make sure I don’t get bellow 30fps. -
I’m running 3 monitors with a AMD 6970 @ 4800×900 and a couple of MFDE monitors on a 5450. While I am only getting normal FPS with all the eye candy in the 30s, (CPU is an AMD [email protected]), I still don’t think the 6970 is the bottleneck. My system monitoring software on one of my MFDE monitors reports that the main GPU rarely gets above 60% load.
-
That is interesting because my hw monitor says the same. But I had also 0-100% inside the MSI Afterburner.
-
Did you heard about the new At drivers that boost performance?
The new AMD Crimson drivers toasted my system…I lost 2 days of flying while trying to get my Pit back to Mission Ready. Admittedly, I tried the beta release, but the final is out there, dtd 17 Dec 2015. I’m skeptical until I see more Crimson versions. Can’t afford the downtime.
-
Lol Yeap that was a big f up. I hope they paid for the replacement.
Now it’s fixed as I read.
For those that don’t know there was an error in the first driver and fan control resulting to melted cards. -
That is interesting because my hw monitor says the same. But I had also 0-100% inside the MSI Afterburner.
I had MSI afterburner but lost it after I had a problem with my new MSI motherboard and had to get a new motherboard and re-install Windows. I’ve been using Moo0, Open Hardware Monitor, or CPUID HWMonitor and they will all report the less than 60% (also when I scale back to 1600x900 they all only report a mid 30% load). I’ll try MSI Afterburner later and see what it reports.
-
I’m not much of a fan of MSI Afterburner now that I’ve used HWiNFO. In MSI, you only get min and max… sure, you can move your mouse over the graph to see what the numbers are at that particular time but that’s about it. HWiNFO has min, max, and avg numbers which gives a better picture of the load on the CPU/GPU. MSI is still good to see a visual representation of the load…
-
Is there some settings that I have missed to get decent fps?
I run a i7 [email protected] with x2 r9 290x’s with x3 1080p screens. With 4.32 I got good fps (around 60fps, with drops with a lot of activity)
Finally got around to playing 4.33 and using the INS Tactical Engagement I am getting less than 15 fps.
Are there any new settings in not aware of that still tank fps?
-
did some more testing, it seems that Crossfire is negatively scaling right now, so I get a 20fps boost disabling crossfire.
-
I’m running 3 monitors with a AMD 6970 @ 4800×900 and a couple of MFDE monitors on a 5450. While I am only getting normal FPS with all the eye candy in the 30s, (CPU is an AMD [email protected]), I still don’t think the 6970 is the bottleneck. My system monitoring software on one of my MFDE monitors reports that the main GPU rarely gets above 60% load.
Hello Snake
I just treated myself to an XFX DD R9 380 4GB to upgrade from an XFX HD 5750 1GB and to honest am astounded at the FPM that I now get (doubled at least from 20-30 FPS to 60+ FPS) on the main monitor at 1920 x 1080 BUT for the MFDs then I have an XFX HD 5450 to drive L & R on 2x mini TFT 800 x 600 resolution using Lightning MFDE to extract. Since I added the R9 380 the MFD monitors driven by the HD 5450 have been running very choppy with the TGP image very laggy/choppy on one MFD and the ‘sweep cursor’ on the FCR also laggy on the other MFD.
I did not yet try BMS in windows extracted mode but I would like to know why the sudden change for the worse for the second graphics card. Any ideas please? -
Hello Snake
I just treated myself to an XFX DD R9 380 4GB to upgrade from an XFX HD 5750 1GB and to honest am astounded at the FPM that I now get (doubled at least from 20-30 FPS to 60+ FPS) on the main monitor at 1920 x 1080 BUT for the MFDs then I have an XFX HD 5450 to drive L & R on 2x mini TFT 800 x 600 resolution using Lightning MFDE to extract. Since I added the R9 380 the MFD monitors driven by the HD 5450 have been running very choppy with the TGP image very laggy/choppy on one MFD and the ‘sweep cursor’ on the FCR also laggy on the other MFD.
I did not yet try BMS in windows extracted mode but I would like to know why the sudden change for the worse for the second graphics card. Any ideas please?I had a similar issue on mine when I switched to a R9 290 with very choppy MFDe, but updating the drivers fixed it for me. https://www.benchmarksims.org/forum/showthread.php?23359-Lightning-s-MFDE-and-4-33/page4 Posts 133-135
Also FWIW on the multi-monitor front, I’ve been playing with PLP setups and while i liked my 20-27-20 setup, I have finally settled on a 20-32-20 4360x1600 which is pretty sweet, I’ll try to get some pictures soon.
-
I had a similar issue on mine when I switched to a R9 290 with very choppy MFDe, but updating the drivers fixed it for me. https://www.benchmarksims.org/forum/showthread.php?23359-Lightning-s-MFDE-and-4-33/page4 Posts 133-135
Also FWIW on the multi-monitor front, I’ve been playing with PLP setups and while i liked my 20-27-20 setup, I have finally settled on a 20-32-20 4360x1600 which is pretty sweet, I’ll try to get some pictures soon.
Thanks for the speedy reply - better to re-install the cards one at a time with Crimson Drivers or leave both in when updating?
-
I had then both installed at the same time for the driver uodate.
-
I had a similar issue on mine when I switched to a R9 290 with very choppy MFDe, but updating the drivers fixed it for me. https://www.benchmarksims.org/forum/showthread.php?23359-Lightning-s-MFDE-and-4-33/page4 Posts 133-135
Also FWIW on the multi-monitor front, I’ve been playing with PLP setups and while i liked my 20-27-20 setup, I have finally settled on a 20-32-20 4360x1600 which is pretty sweet, I’ll try to get some pictures soon.
The drivers were the culprit Snake - updated with the latest 16.1 Crimson drivers and now running fine thanks.
I still drop from 60-70 FPS to around 20 FPS with TGP and MAV on 7" LCD MFD screens using MFDE but at least the choppiness has gone. I think that the CPU could be bottlenecking a bit, but BMS is still playable without stutter. Not the huge improvement that I was hoping for from my R9 380 investment but at least I partially ‘future proofed’ my PC and I will stop chasing the FPS Nirvana -
Good to hear! I too saw almost no FPS improvement at 4800x900 with the upgrade to a 290, but I wonder now what 4860x1600 would have looked like on the 6970, I bet it would have been the difference. I think Falcon is always CPU bottlenecked.
Tip on the TGP, there is a point around 20000’ you get a FPS jump and the TGP hit on my system isn’t that bad above the altitude of that jump.
-
well going from 900 to 1600 will have an enormous impact.
All VGA’s suffer mostly on the vertical resolution.
-
Interesting Arty. I compared it by plain old pixel math when trying to estimate performance.
For reference the triple monitor resolutions I’ve played with:
4800x900=4,320,000 (3x20 16:9s)
3720x1600=5,952,000 (20-27-20 PLP, the 20s in portrait mode and native resolution, this causes dot pitch issues, not used much)
3720x1440=5,356,800 (20-27-20 PLP, 20s resolution turned down to 1440x900 to reduce the difference in the vertical of the dot pitch)
4860x1600=7,680,000 (20-32-20 PLP, perfect dot pitch match)
Also for reference 1920x1080 HD is 2,073,600 and 4K is 8,294,400.Part of the problem here is that the 4800x900 was a standard Eyefinity supported setup, I was running with V-Sync on, and was the only resolution ran on the 6970. I have found now that at least with SoftTH either SoftTH, AMD, or Falcon’s way it fills the V-Sync buffer actually causes a decent FPS hit if you are consistently running under 60 FPS, so unfortunately I can’t directly compare what it would have been now but then I was getting in the high 30s to low 40 most general flying around under 20,000’ with all the eye candy on/max. Now I am curious if I had turned off the V-Sync how much more I would have had since I was only pushing 4,320,000 pixels I bet it would have been higher. But again the difference between the HD 6970 and the R9 290 was negligible and they were an “apples to apples” comparison. In synthetic benchmarks say the 290 is about twice as powerful, which of course never equates to a 100% increase in game, but just throwing it out there.
The comparison problem with the PLP resolutions is that I only ran them on the 290, with SoftTH, and found out V-Sync wasn’t helping in these setups. So these are the “oranges to oranges.” 3720x1440 or 3720x1600 while having 10% difference (that is both in the total pixels or vertical resolution) did not result in a 10% hit. They were practically the same, mid to high 30s FPS. Going from them to 4860x1600 is about a 40% jump in pixels, and I lost only about 15% of FPS and was able to get the hit down to under 10% by turning off the shadow details. So either you are correct that it is mostly vertical resolution that makes the difference AND/OR, my theory that BMS is still largely CPU bound since I’m throwing a much bigger graphical load at it and the results in the same general performance. I’m also running about 93% of 4K in pixels but only 74% of it’s vertical resolution. That might be a good way to test to see if it is more the number of pixels or the vertical resolution has a bigger effect, but I don’t have access to a 4K display and it isn’t a supported VSR oversampling resolution.
3x1 20" 4800x900
PLP 20-27-20, 3270x1440. It was always a interim setup in my book due the dot pitch differences and wanting a wider hFOV. Most PLP guys don’t use the “hang” (bottom part of the monitor that goes past the bottom of the main monitor), but I’ve found that I kind like it in BMS because it kind simulates the visibility you get around the sides of the cockpit.
PLP 20-32-20 4860x1600, still a work in progress with the peripherals and monitor positioning. The hang here feels and is a little more away from the sides of the cockpit panel and is overall not as big. In the future I may get stop using it and just run 4860x1440.In general PLP has been a positive to me, you still get most of the wider hFOV and a better vFOV while pushing out the bezels further to the side of your vision that makes them less noticeable. Also switched out the instrument monitors to 20" 16:9 FWIW.
-
Oh you are getting in to waters where resolution and plp or lll or ppp or vsynch is not only the issue.
bw of vram and bit 256? 384?
bw of the bus. this has to do with mobo
bw with mobo ram and CPU
bw of CPU and pcie…all those must be in perfect or near perfect synch to get the most out of your hardware on a flight sim not just falcon.
specially when the pixels you demand to show up are that high, and specially in the vertical. -
This post is deleted!