May 14, 201313 yr Quick summary: 10-15% (2/2.5 to 5 fps) gain for 500mhz increase. Read on if you want details So normally I run what I guess you would call an "adaptive overclock". My i7-860 normally runs at 2.8Ghz. I have it set at 150x21 with speedstep and C-States on. This way it normally runs at 3.3Ghz (speedstep applies a 22 mult. when all cores in use instead of 21) and as cores go unused and are shut off it could reach 3.9Ghz on one core (not that it ever gets there). Typically if I'm in the KSEA area in FSX with the OrbX scenery there, cpu-z shows I'm running solid at 3.3Ghz and not really anything more. Typically I get 24-40fps flying between KSEA and downtown at about 2200ft in the default A320 zoomed all the way out in spot view. When I'm at altitude I like to run in spot view as much as possible for "sightseeing" so this is the way I test FPS. I also get around 28-35 FPS at KSEA itself. Now, in an effort to try and get a few more FPS without turning down settings (not really sure what settings would have the most effect, probably autogen which is on normal right now) I've been working on a new OC. I turned off HT since FSX doesn't use it as far as I understand so I could save on heat since the OC would only be used for FSX. I went for 180x21 (i.e. 3.8Ghz basically). Now to the question. I gained about 10-15%, generally closer to 10-12% increase in FPS, and maybe a slight bit of smoothness. Do you guys think that is worth it for the extra 500mhz? The actual fps gains are around 2/2.5-5fps depending if you're looking at the high or low end of the original ranges. I know it doesn't sound like a lot, but it IS 10% or more. I might see if I can hit 4.2ghz with 200x21, but I already know I either can't go above about 184 or I have a "bclock hole" around 184-19x. I was trying 190 with the multiplier way low to isolate the memory controller from the cpu and between 180-190 it seemed like it was going to want a .05v increase in vtt to maybe be stable (never tested it, it failed still after the first increase of .025v). So it seems like 4ghz might be out, but if I'm just in a stability "hole" around 190 then maybe I just need to give 4.2 a shot and see if I'm on the other side of it. AMD Ryzen 9950X3D | Asrock X870E Taichi | Gigabyte Gaming OC 4090 w/EK waterblock | Full Custom Loop Cooling | Corsair Dominator Titanium DDR5-6000 | Samsung & WD NVME/SSDs | Phanteks Enthoo 719 | Seasonic Vertex Gold 1200W | Keychron Q5 Max | Corsiar Scimitar Elite SE Wireless | Honeycomb Alpha and Bravo | Logitech Pro Flight Pedals | VKB Gladiator Pro NXT L&R handed | MiniCockpit MiniFCU | Alienware AW34DWF | Asus PG279Q | Win 11 Pro
May 14, 201313 yr How much hotter does it get? I know 2fps gain to me is noticeable. But its more about the smoothness, IE you might get 28fps but occasional spikes to lower numbers which OCing can sometimes stop, this is even more noticeable. James Bennett
May 14, 201313 yr Author I think FSX itself was running maybe 1-2C hotter (sounds low but remember I turned off HT). I don't know as far as something like Intel Burn Test which I've been running at each step of the process because it was late last night when I finished the bclock and memory testing and I just decided to ramp up the cpu to "full speed ahead" and see if it would boot/make it into FSX. Since it did I went ahead and checked FPS so I could get a feel for if it would be worth it to finish the process and test/adjust for stable CPU overclock & voltage. Sounds like it may be worth it to go ahead and do that. Oh I did forget to mention I normally run externally locked to 30fps in Inspector with 1/2 refresh rate vsync forced. I just removed those two options for fps testing of course. Oh and of course I changed to the appropriate affinity mask for the overclock since I turned HT off. AMD Ryzen 9950X3D | Asrock X870E Taichi | Gigabyte Gaming OC 4090 w/EK waterblock | Full Custom Loop Cooling | Corsair Dominator Titanium DDR5-6000 | Samsung & WD NVME/SSDs | Phanteks Enthoo 719 | Seasonic Vertex Gold 1200W | Keychron Q5 Max | Corsiar Scimitar Elite SE Wireless | Honeycomb Alpha and Bravo | Logitech Pro Flight Pedals | VKB Gladiator Pro NXT L&R handed | MiniCockpit MiniFCU | Alienware AW34DWF | Asus PG279Q | Win 11 Pro
May 15, 201313 yr Author Ok just finished my OC testing. I had to drop from 3.78ghz to 3.7 because I could not get the Intel Burn Test tool to even pass 3 rounds even after bumping up .05v vcore. However I think the results are good. probably 8/9-12% increase in my FPS tests over Orbx Seattle area. It did look like low end spikes got reduced making things smoother too. So, looks like I have a special overclock just for FSX. This weekend I may get crazy and try to work around what I think is a bclock hole and go for 4.2Ghz but I don't have high hopes for that. So, overall compared to my normal overclock that varies from 3.3-3.9ghz with speedstep I have the same vcore, one notch higher vtt, and probably a few degrees higher temp. Not bad for a solid 3.7ghz clock (keeping in mind HT is disabled on it). AMD Ryzen 9950X3D | Asrock X870E Taichi | Gigabyte Gaming OC 4090 w/EK waterblock | Full Custom Loop Cooling | Corsair Dominator Titanium DDR5-6000 | Samsung & WD NVME/SSDs | Phanteks Enthoo 719 | Seasonic Vertex Gold 1200W | Keychron Q5 Max | Corsiar Scimitar Elite SE Wireless | Honeycomb Alpha and Bravo | Logitech Pro Flight Pedals | VKB Gladiator Pro NXT L&R handed | MiniCockpit MiniFCU | Alienware AW34DWF | Asus PG279Q | Win 11 Pro
Create an account or sign in to comment