I think Gearjammer is asking you to see how many FPS you get at 800x600 with you system not overclocked. To be compared to the results you posted above. Perhaps this would be a more direct rF2 overclocking benefit comparison than 3DMark scores.
You are correct Ricknau, I want to see the fps running at the low res with both OC and no OC to see how much difference there is in fps. This will show how much benefit there is to doing the actual OC. Also try and do the fps test with a replay so that we know the conditions don't change on what is displayed. As to the OP's question, I still say that you will not benefit nearly as much with a CPU swap as you would with a GPU swap. With the GPU swap you may well see up to a 20% increase in FPS where the CPU is likely to be limited to under 10% gain. The reason is your CPU isn't really a bottleneck or shouldn't be anyways.
Sorry Doh idiot. lopl Here is the other 2 I re-ran at max turbo speed. 3.7GHz ===== 800x600...........116FPS 1920 x 1080...... 88FPS 4.5GHz ===== 800x600...........151FPS 1920 x 1080...... 106FPS It is only a rough test but was not much variation tried each 3 times or so. Overclock used more vram at both resolutions, interesting.
Sorry, i don't want to intrude But imho this kind of tests only make sense, if all mainly gpu related eyecandy are also turned off.
Why when you are never going to do that, makes even less sense to me. No use buying a processor or GPU based on anything it can do less then your preferred resolution and quality.
Because mainly gpu related tasks have hardly anything to do with the cpu, and will first of all disort the results. If a cpu could run the game with (for example) 100fps at low resolution and without all mainly gpu related eyecandy. The same cpu should be able to run nearly the same fps at higher resolution and eyecandy on, as long as the gpu could manage it.
I agree a GPU is the first to look at but where does he go from a 7970 ? That is a dear card he has to sell and replace, you would want to replace it with at least a GTX770 but that would not be good bang for buck let alone a GTX780 because I don't see his current CPU at stock clock speed pushing either of those cards to their potential. Catch22
I understand why sheesh lol In the real world it does not matter though. edit: If you have new system with high end CPU /GPU you baseline test at your resolution and settings at stock clocks and overclocks. Anything else is meaningless really.
I beg to differ. Most player here have enough fps as long as they reduce their settings. So if the chief target is to get more eyecandy without losing too much fps, a better gpu is the first choice. Of course, a faster cpu can't hurt, especially with a dual gpu or sli this will be essential. But thats not the case here. thats all ...
I just did a GPU swap. I got a 7970. I did see a fps improvement, but not nearly what I was hoping for. So will I get more out of a new CPU with a better GPU, or are they totally independent where fps are concerned?
Well it's hard for me to figure what is the more meaningful test but looking a DD's numbers with his CPU OC'd by ~22%... With a GPU coasting at low res the FPS goes up by ~30%. With a GPU reasonably taxed though (maybe at/near its limit?) DD gets a ~21% FPS gain. I don't want more eye candy, I just want more FPS with the candy I have now. My 560ti with 1.28MB is at it's limit now. I'm just wanting to convince myself if it will be worth my effort to OC my CPU (and then my GPU) with the card I have. Maybe forgo a new graphics card at least for a while. Actually, as I think about it, it would seem impossible to not have gains proportional to the clocking increase, especially if both processors are sped up. I just need to get up the gumption to pursue it. I think I'll start shopping for a good CPU fan/heatsink.
Capriciously logic! If you think again about it - it's both the same. I was in a equal situation. With my old 560i i was able to play with highest settings, but the lack of fps made it almost unplayable. On the other side i reduced my settings and got playable fps, but with the lack of eye candy. What does that tell you? - Both things are inextricably!
Thanks for those results DD, it clearly shows that an OC'd CPU does in fact have a decent increase with your graphic settings left alone. What the other poster was talking about though is if the GPU were still being held back at lower res with the graphics at maxed, by reducing the graphics you would be sure there was nothing there to alter the outcome. In a nutshell, by increasing your speeds and voltages you were able to obtain at least a 20% increase in fps over stock, which is good. This shows that there are still things happening in rF2 that are CPU dependent and will benefit from an increase in the clocks. So to answer ccjcc81's question, yes you may well see an increase as long as the stock clock of the new CPU is appreciably higher than the one you are replacing. If you plan on overclocking, then you would benefit more from an i5 over the AMD chips as you are likely to achieve faster clocks with the Intel processor. If you only plan on running stock clocks, then no, you will not benefit from a CPU swap.
Hey that worked pretty good. Used Gigabyte "Quick Boost" to get to 4.3GHz with a click of the mouse. I won't test under load until I get better cooling but at least it boots and runs Windows. That's better than my previous luck. Well no sh!t. I'm just saying my trade-off will be for more FPS, not eye candy. If I can get my machine turboed up that will be my payoff. "Capriciously logic?" I need to look that one up. Perhaps a OC thread is in order. I have questions if anyone wants to continue down that path with me.
You can test under load but you want to have a cool room and blow a fan inside the case If you keep it below 70C is no problem just to test for stabilty in 3D I would no worry about stress tests too much till you get a cooler. Which board do you have that is good overclock barely anything between 4.3-4.5 Auto overclock usually gives 4.2Ghz -4.5GHz What you want to do is check settings in bios how it set then go manual again and just drop voltage a notch at a time On some you can do 4.3GHz at near stock voltages. OC Genie tend to set voltage higher, lower you can get with no BSOD the better. Also make sure other voltages are stock You do not need to puch other voltages other then CPU to 4.5GHz you can also disable the auto power load thingie you do not need that till about 4.8GHz Then if you disable all power settings Disable turbo See how it runs a loop of 3dmark or something , if it does not get too hot that is. You can see if timings and voltages are solid if it gets through any 3dmark version. I do not even worry about stress tests these days it is pointless with these chips they just run too good. Get "HWinfo" ( OSD sensors) , set to run sensors, same time as tests so you can see core temps. Run rF2 offline @Turbo then again @ 4.3GHz
That was low stress track too, look at vram usage. Silverstone would do around the same variation you think ?
Only a few Examples how (un)important the CPU in Full HD and higher Resolution is All Systems with: 8Gb RAM, AMD HD 7970, 128 GB SSD
Well sorry but it is un important Look if people do not think that do not preach to me. If I buy a HIGH END CPU and GPU I am not bloody interested what it does at 1280 WHO CARES !! All I care about is what rFactor2 runs like at 1920 or 2560 or whatever you run. YES YES YES YES YES for people that cant manage a steady rate maybe it is interesting, don't see how it can help. Are you seriously going to tell me you would run any of those games under 1920 to get more fps. ? No you would drop settings, shadows and AA first ! oh dear lol If you have a slower performing PC disregard what I say but we were talking about a IVY and 7970 min.
Wasn't it more usefull rf2 would use all cores instead to overclock just because rf2 doesn't utilize all cores a CPU has ? Sorry if it is stupid but i really don't understand why is that. Imagine half of the zyl. of a V6-V8-V12 or whatever engine don't get fuel. Why i should tune the other half ? Not very logical to me. Please upgrade the rf2 Engine to utilize all available cores.
The point here is that as the resolution increases, the GPU becomes the limiting factor. With low resolution the GPU has no problem producing frames and the CPU cannot keep up. I find DD's results very impressive and the OC is certainly working for him. On my system with i5 3570k, GTX770, 16GB ram and running 2560 by 1440 with everything max; going from 3.8 to 4.4GHz yeilds only 5-10% FPS increase. Also, this increase only seems to occur on portions of the track where FPS were high to begin with. On the starting grid and anywhere on track where FPS normally gets lower, the OC does not help. Back to the OP, if you are to upgrade, 100% go with intel 4th gen and take advantage of a better performing mobo and system performance altogether.