- Messages
- 66
- Country

Hi all,
Long time since I've posted here, but I know the experts read posts here, and my discussion is probably quite technical.
I recently got back into simming, first because I inherited a slightly older (but capable) PC, then I had the opportunity to move to a pretty much top-line PC in the last month or two.
Here's the two setups:
BOX A (2010):
I7-950 overclocked to 4.1ghz, GTX 970 w/4gb, 8 gb DDR3 ram
BOX B (2017):
I7-7700k overclocked to 4.7ghz, GTX 1080 w/8gb, 32 gb DDR4 ram
In terms of benchmarks, the newer machine (B) is a magnitude ahead of the older.
Both are running win10x64 pro, and I've installed FSXse with identical addons that include OrbX global, vector, LC, and some airports and regions, Active Sky 2016 and REX texture direct,default and addon (A2A, Carenado, others). Descriptions below are all with Cessna 172/182, no heavy metal which bring some additional burden...
I've also tweaked both similarly with FSX.cfg settings (buffer pools on the GPU, lod at 5.5 or less, highmem fix usually on, though I've tried off, tested various affinity masks, that's it) and I'm running DX10 with steve's fixer.
What I've observed:
1. Outside of major hubs (NYC, LA, etc.) performance is excellent on both machines, but noticeably better on the newer (box b) - probably averaging 30-50+ FPS on new machine, 20-30+ FPS on older (box a). In same areas, textures are sharper on new machine than old, even though settings are the same. I also have the sense that more 3D objects load on new machine even though both scenery and autogen density settings are the same.
2. In dense urban areas and major hubs performance is better on newer machine (box b) with many autogen buildings loading than on older (box a). FPS on newer machine averages around 20, older 15 but can drop to single digits. However CTDs due to memory errors are now much more prevalent on the new machine and RARE (in fact, almost absent) on the old machine. My sense is that the new machine is able to process more 3D scenery and rapidly consumes VAS, whereas the older box is able to throttle itself better and conserves VAS.
My question for the experts -
* Anyone else experience this?
* Any merit behind my hypothesis?
* Anything to do about it?
On the same machines, XP11 runs really smooth. I think predictably better on the new machine vs. older. If the ground handling, weather and AI/ATC improve, along with some a few higher fidelity aircraft, XP will really be a competitor.
Long time since I've posted here, but I know the experts read posts here, and my discussion is probably quite technical.
I recently got back into simming, first because I inherited a slightly older (but capable) PC, then I had the opportunity to move to a pretty much top-line PC in the last month or two.
Here's the two setups:
BOX A (2010):
I7-950 overclocked to 4.1ghz, GTX 970 w/4gb, 8 gb DDR3 ram
BOX B (2017):
I7-7700k overclocked to 4.7ghz, GTX 1080 w/8gb, 32 gb DDR4 ram
In terms of benchmarks, the newer machine (B) is a magnitude ahead of the older.
Both are running win10x64 pro, and I've installed FSXse with identical addons that include OrbX global, vector, LC, and some airports and regions, Active Sky 2016 and REX texture direct,default and addon (A2A, Carenado, others). Descriptions below are all with Cessna 172/182, no heavy metal which bring some additional burden...
I've also tweaked both similarly with FSX.cfg settings (buffer pools on the GPU, lod at 5.5 or less, highmem fix usually on, though I've tried off, tested various affinity masks, that's it) and I'm running DX10 with steve's fixer.
What I've observed:
1. Outside of major hubs (NYC, LA, etc.) performance is excellent on both machines, but noticeably better on the newer (box b) - probably averaging 30-50+ FPS on new machine, 20-30+ FPS on older (box a). In same areas, textures are sharper on new machine than old, even though settings are the same. I also have the sense that more 3D objects load on new machine even though both scenery and autogen density settings are the same.
2. In dense urban areas and major hubs performance is better on newer machine (box b) with many autogen buildings loading than on older (box a). FPS on newer machine averages around 20, older 15 but can drop to single digits. However CTDs due to memory errors are now much more prevalent on the new machine and RARE (in fact, almost absent) on the old machine. My sense is that the new machine is able to process more 3D scenery and rapidly consumes VAS, whereas the older box is able to throttle itself better and conserves VAS.
My question for the experts -
* Anyone else experience this?
* Any merit behind my hypothesis?
* Anything to do about it?
On the same machines, XP11 runs really smooth. I think predictably better on the new machine vs. older. If the ground handling, weather and AI/ATC improve, along with some a few higher fidelity aircraft, XP will really be a competitor.





