As promised I will be providing a quick summary of my in-game Arma 2 experiences and more general experiences with my new Sandy Bridge setup (thank you for the mess up OcUK 
). I haven't had a chance to test much but I am rather impressed with myself because everything worked flawlessly the second I powered the system on. I haven't had massive amounts of experience building systems but I seemed to have learnt the art quite well.
The first thing to mention is, of course, the changes that have occured in my system:
- MSI NEO 2 P35 motherboard --> ASUS P8P67
- 4GB G.Skill RAM @ 1000Mhz effective --> 4GB G.Skill Ripjaw 1680Mhz effective
- Intel Q6600 G0 @ 3.2Ghz --> Intel Core i5 2500k @ 4.3Ghz Turbo
- Arctic Freezer 7 Pro Revision 2
Other aspects of my system that may be of interest to you are the GPU (Radeon 5850 @ around 5870 clock speeds) and the presence of a dedicated Xonar D2X PCI-E soundcard. The system also uses X25M 80GB and Vertex 2E 60GB SSDs.
Before you gasp in disbelief at the above overclock thinking my i5 2500k should be capable of more - you are probably correct and it probably is. The 4.3Ghz turbo was achieved completely automatically using the ASUS 'TurboV Evo Auto Tuning' feature which, as you can see, worked very nicely. It also raised the RAM speed slightly from 1600Mhz --> 1680Mhz without upsetting the timings or voltages. The CPU voltage was also untouched by the utility if not lowered slightly for the hell of it to 1.195V (I think - unless this is just an idle voltage).
Here are a few pictures of the new kit all put together (before becoming a full system, of course):
http://www.pcmonitors.org/wp-content/uploads/2011/01/SB-side.jpg
http://www.pcmonitors.org/wp-content/uploads/2011/01/SB-top.jpg
Now onto the all-important Arma 2 experiences. Bear in mind that this testing is not supposed to be a particularly scientific or precise comparison of the two processors as the RAM (in both cases 4GB) and motherboard is of course different and at times the single overclocked 5850 will be a limiting factor during the strenous benchmarking (settings below but note brightness was temporarily increased to rectify a problem with an overly dark modded nightvision device. During actually testing it was set to 1.2):
This was a particular game title where, during actual gameplay, the Q6600 @3.2Ghz presented a bottleneck. This bottleneck was evident using the same graphics settings as the benchmark (view distance anywhere above around 3500m) during normal gameplay on the Q6600-based system. In heavily built up urban areas and the hills overlooking them, in particular, the frame rate would dip below 25fps. This could not be alleviated by reducing the 3D Resolution as low as it would go, suggesting the GPU was not presenting a bottleneck here.
The performance in the places that proved problematic on the Q6600 was much, much better with the 2500k. Instead of being still in the doldrums at a miserable 20-25fps the frames were soaring away at 45fps+.I have tried to capture some of the differences in a quantitative and comparative sense using some of the in-game benchmark scenarios. There is something a bit odd about the way the benchmarks operate as there seem to be certain limitations (in minimum framerate in particular) that simply don't exist if you revisit the benchmark areas in game. Nonetheless here are the results:
Operation Arrowhead benchmark:
This benchmark pans around a dense village area populated by some gun-bearing locals being attacked by US forces. The minimum framerates seemed to occur in a fairly dense area with a group of strange engine limitation in the benchmark at this point - as aforementioned if you visit this place in the game the frame rate is much higher (generally around 40-50fps with the 2500k set at 4000-6000m view distance).
View Distance = ~6500m
Q6600 @3.2Ghz
Min= 18
Avg= 35
Max= 54
2500k @4.3Ghz
Min= 26
Avg= 43
Max= 63
----------------
View Distance = ~3250m
Q6600 @3.2Ghz
Min= 17
Avg= 37
Max= 62
2500k @4.3Ghz
Min= 26
Avg= 47
Max= 71
Arma 2 benchmark 1, day:
This benchmark pans around a heavily forested area at the start of the benchmark with soldiers infront of trees, camera zoomed up close and panning around. The engine doesn't seem to like this panning near these trees but the framerates easy off a bit further through the benchmark.
View Distance = ~6500m
Q6600 @3.2Ghz
Min= 15
Avg= 23
Max= 51
2500k @4.3Ghz
Min= 31
Avg= 41
Max= 79
----------------
View Distance = ~3250m
Q6600 @3.2Ghz
Min= 21
Avg= 32
Max= 62
2500k @4.3Ghz
Min= 32
Avg= 47
Max= 78
Arma 2 benchmark 2, night fight:
This benchmark contains exceptionally heavy volumes of tracer fire (unlike you're ever likely to see in the actual game) and many simultaneous explosions during a night scene. Very strenous stuff.
View Distance = ~6500m
Q6600 @3.2Ghz
Min= 6
Avg= 11
Max= 18
2500k @4.3Ghz
Min= 14
Avg= 21
Max= 24
----------------
View Distance = ~3250m
Q6600 @3.2Ghz
Min= 7
Avg= 13
Max= 18
2500k @4.3Ghz
Min= 18
Avg= 24
Max= 26
Make what you will of these benchmark results but I can honestly say this has been a worthwhile upgrade for Arma 2. I haven't really tested much else out yet but I am happy with the new system. Also as a closing note - the EFI bios is very nice indeed and the system does boot up slightly faster. GPU will be the next upgrade no doubt but I will probably be waiting for the next generation of those.
). I haven't had a chance to test much but I am rather impressed with myself because everything worked flawlessly the second I powered the system on. I haven't had massive amounts of experience building systems but I seemed to have learnt the art quite well.The first thing to mention is, of course, the changes that have occured in my system:
- MSI NEO 2 P35 motherboard --> ASUS P8P67
- 4GB G.Skill RAM @ 1000Mhz effective --> 4GB G.Skill Ripjaw 1680Mhz effective
- Intel Q6600 G0 @ 3.2Ghz --> Intel Core i5 2500k @ 4.3Ghz Turbo
- Arctic Freezer 7 Pro Revision 2
Other aspects of my system that may be of interest to you are the GPU (Radeon 5850 @ around 5870 clock speeds) and the presence of a dedicated Xonar D2X PCI-E soundcard. The system also uses X25M 80GB and Vertex 2E 60GB SSDs.
Before you gasp in disbelief at the above overclock thinking my i5 2500k should be capable of more - you are probably correct and it probably is. The 4.3Ghz turbo was achieved completely automatically using the ASUS 'TurboV Evo Auto Tuning' feature which, as you can see, worked very nicely. It also raised the RAM speed slightly from 1600Mhz --> 1680Mhz without upsetting the timings or voltages. The CPU voltage was also untouched by the utility if not lowered slightly for the hell of it to 1.195V (I think - unless this is just an idle voltage).
Here are a few pictures of the new kit all put together (before becoming a full system, of course):
http://www.pcmonitors.org/wp-content/uploads/2011/01/SB-side.jpg
http://www.pcmonitors.org/wp-content/uploads/2011/01/SB-top.jpg
Now onto the all-important Arma 2 experiences. Bear in mind that this testing is not supposed to be a particularly scientific or precise comparison of the two processors as the RAM (in both cases 4GB) and motherboard is of course different and at times the single overclocked 5850 will be a limiting factor during the strenous benchmarking (settings below but note brightness was temporarily increased to rectify a problem with an overly dark modded nightvision device. During actually testing it was set to 1.2):
This was a particular game title where, during actual gameplay, the Q6600 @3.2Ghz presented a bottleneck. This bottleneck was evident using the same graphics settings as the benchmark (view distance anywhere above around 3500m) during normal gameplay on the Q6600-based system. In heavily built up urban areas and the hills overlooking them, in particular, the frame rate would dip below 25fps. This could not be alleviated by reducing the 3D Resolution as low as it would go, suggesting the GPU was not presenting a bottleneck here.
The performance in the places that proved problematic on the Q6600 was much, much better with the 2500k. Instead of being still in the doldrums at a miserable 20-25fps the frames were soaring away at 45fps+.I have tried to capture some of the differences in a quantitative and comparative sense using some of the in-game benchmark scenarios. There is something a bit odd about the way the benchmarks operate as there seem to be certain limitations (in minimum framerate in particular) that simply don't exist if you revisit the benchmark areas in game. Nonetheless here are the results:
Operation Arrowhead benchmark:
This benchmark pans around a dense village area populated by some gun-bearing locals being attacked by US forces. The minimum framerates seemed to occur in a fairly dense area with a group of strange engine limitation in the benchmark at this point - as aforementioned if you visit this place in the game the frame rate is much higher (generally around 40-50fps with the 2500k set at 4000-6000m view distance).
View Distance = ~6500m
Q6600 @3.2Ghz
Min= 18
Avg= 35
Max= 54
2500k @4.3Ghz
Min= 26
Avg= 43
Max= 63
----------------
View Distance = ~3250m
Q6600 @3.2Ghz
Min= 17
Avg= 37
Max= 62
2500k @4.3Ghz
Min= 26
Avg= 47
Max= 71
Arma 2 benchmark 1, day:
This benchmark pans around a heavily forested area at the start of the benchmark with soldiers infront of trees, camera zoomed up close and panning around. The engine doesn't seem to like this panning near these trees but the framerates easy off a bit further through the benchmark.
View Distance = ~6500m
Q6600 @3.2Ghz
Min= 15
Avg= 23
Max= 51
2500k @4.3Ghz
Min= 31
Avg= 41
Max= 79
----------------
View Distance = ~3250m
Q6600 @3.2Ghz
Min= 21
Avg= 32
Max= 62
2500k @4.3Ghz
Min= 32
Avg= 47
Max= 78
Arma 2 benchmark 2, night fight:
This benchmark contains exceptionally heavy volumes of tracer fire (unlike you're ever likely to see in the actual game) and many simultaneous explosions during a night scene. Very strenous stuff.
View Distance = ~6500m
Q6600 @3.2Ghz
Min= 6
Avg= 11
Max= 18
2500k @4.3Ghz
Min= 14
Avg= 21
Max= 24
----------------
View Distance = ~3250m
Q6600 @3.2Ghz
Min= 7
Avg= 13
Max= 18
2500k @4.3Ghz
Min= 18
Avg= 24
Max= 26
Make what you will of these benchmark results but I can honestly say this has been a worthwhile upgrade for Arma 2. I haven't really tested much else out yet but I am happy with the new system. Also as a closing note - the EFI bios is very nice indeed and the system does boot up slightly faster. GPU will be the next upgrade no doubt but I will probably be waiting for the next generation of those.
Last edited:
