System performance of version 3.0

In this thread I will post our findings about frame rates.

A few words first about the methodology: We created a benchmark scenario testing three different scenes,

1. Open range, some trees, a distant city, external observer's position
2. Commander's position unbuttoned, urban scene, with dismounted soldiers and a few civilians roaming the streets
3. Same scene, but as seen through a gunner's sight

Of each scene, we took a sample of the frame rate, and gave equal weighting to these frame rates to build an average. Typically the weapon sight offers the highest frame rate, the commander's unbuttoned view the lowest.

Most tests were made with screen resolutions "Full HD" (=1920 x 1080) and "HD" (=1280 x 720).

We will use six categories to grade performance,

1. Unusable:
   Frame rates are too low to operate the simulation effectively.
2. Not recommended:
   Either the average frame rate, or just a single one of the three scenes yields a frame rate of under 24 fps (no rounding up allowed - for this, or all other categories)
3. Low performance:
   Both the average and the minimum frame of all three scenes rate must be above 24 fps
4. Medium performance:
   The average frame rate must be above 36, the lowest frame rate above 30 fps
5. Good performance:
   The average frame rate must be above 48, the lowest frame rate above 40 fps
6. Great performance:
   Both minimum and average frame rate must exceed 55 frames per second

Version 3.0 will offer a number of slider settings for the visual quality. We have found that most of them have only a very minuscule effect on the frame rates (like .3 ... 1.5 frames per second), except for a) the particle effects, which I cannot discuss yet, and b) the shadow mapping settings.

Shadow mapping will be made available in six level settings,

• Level 0:
  "No shadows" - like version 2.654 and older
• Level 1:
  "Self-shadowing" - tank turrets and gun barrels will cast a shadow on the hull, but no shadows on the ground
• Level 2:
  "Real" shadows - of hull and turret, on each other, and on the ground. Also for buildings.
• Level 3...5:
  Gradually intensification of the shadow, and a higher resolution of the shadow map (finer detail); levels 4 and 5 are recommended only for graphics cards with more than 1 GByte of video RAM

Because the other settings - to the extent that we can tell at this point - are of little to no practical relevance, I will focus on the effect of Shadow Mapping alone (and of screen resolutions, and CPU speed).

The benchmark that we created is a platoon...company level exercise. The village is populated by about 100 civilians, there's about a company worth of armored and mechanized units on the Blue side of it, and no enemy. If you look at scenarios involving a lot more units - like, a reinforced battalion of Blue forces confronting a depleted regiment of Red forces, plus a few hundred refugees in cars and on foot, the graphics card is no longer the bottleneck, but the CPU will become one. Therefore, the reported frame rates have to be seen in this light.

Other findings:

SLI/Crossfire setups do not seem to offer an advantage. This may be due to the fact that in the test setup the CPU was the bottleneck, so on other systems this may not necessarily be the case.

For most graphics cards, reducing the screen resolution to less than 1280 x 720 offered no substantial gain in performance. In other words, HD resolution with at least 2x FSAA should come as a "free" setting; for most, 4x antialiasing will not be felt in the frame rate either.

Top tier (Shadow Mapping Level 5):

Systems that could run Shadow Mapping level 5:

1. GeForce GTX 780, i-7 4770K @ 3.4 GHz, 2560 x 1440: Good performance
2. GeForce GTX 780, i-7 4770K @ 3.4 GHz, Full HD: Great performance
3. GeForce GTX 670, i-7 2600 @ 3.4 GHz, Full HD: Good performance (speculation)
4. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, HD: Medium performance
5. GeForce GTX 580, i-7 2600 @ 3.4 GHz, Full HD: Not recommended
6. GeForce GTX 560 Ti, Phenom-II @ 3.31 GHz, Full HD: Not recommended
7. Radeon HD 57x0, i-5 750 @ 2.67 GHz, Full HD: Not recommended
8. Radeon HD 5970, E8600 @ 3.33 GHz, HD: Not recommended
9. Intel HD4000, i-7 3720 @ 2.6 GHz, HD: Unusable

Systems that could run Shadow Mapping level 4:

1. GeForce GTX 780, i-7 4770K @ 3.4 GHz, Full HD: Great performance
2. GeForce GTX 670, i-7 2600 @ 3.4 GHz, Full HD: Great performance (speculation)
3. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, HD: Medium performance
4. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, Full HD: Low performance
5. Radeon HD 7850, i-7 940 @ 2.93 GHz, HD: Low performance
6. GeForce GTX 650, E8600 @ 3.33 GHz, HD: Not recommended
7. Radeon HD7850 X-Fire, i-7 940 @ 2.93 GHz, Full HD: Not recommended
8. Radeon HD 57x0, i-5 750 @ 2.67 GHz, Full HD: Not recommended
9. Intel HD4000, i-7 3720 @ 2.6 GHz, HD: Unusable

Systems that could run Shadow Mapping level 3:

1. GeForce GTX 780, i-7 4770K @ 3.4 GHz, Full HD: Great performance
2. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, Full HD: Medium performance
3. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, HD: Medium performance
4. GeForce GTX 580, i-7 2600 @ 3.4 GHz, HD: Low performance (estimated)
5. Radeon HD 4870, E8600 @ 3.33 GHz, HD: Not recommended
6. Radeon HD 57x0, i-5 750 @ 2.67 GHz, Full HD: Not recommended
7. Intel HD4000, i-7 3720 @ 2.6 GHz, HD: Unusable

Systems that could run Shadow Mapping level 2:

1. GeForce GTX 780, i-7 4770K @ 3.4 GHz, Full HD: Great performance
2. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, HD: Good performance
3. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, Full HD: Medium performance
4. GeForce GTX 560 Ti, Phenom II @ 3.31 GHz, HD: Low performance
5. Radeon HD 6850, E8400 @ 3.0 GHz, HD: Low perfomance
6. Radeon HD 5730, E8500 @ 3.17 GHz, HD: Low performance
7. Radeon HD 6850, E8400 @ 3.0 GHz, Full HD: Not recommended
8. Radeon HD 5730, E8500 @ 3.17 GHz, Full HD: Not recommended
9. Radeon HD 57x0, i-5 750 @ 2.67 GHz, Full HD: Not recommended
10. Radeon HD 4870, E8600 @ 3.33 GHz, HD: Not recommended
11. Intel HD4000, i-7 3720 @ 2.6 GHz, HD: Not recommended (with no antialiasing activated, otherwise unusable)

Systems that could run Shadow Mapping level 1:

1. GeForce GTX 780, i-7 4770K @ 3.4 GHz, Full HD: Great performance
2. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, Full HD: Great performance
3. Radeon HD 57x0, i-5 750 @ 2.67 GHz, Full HD: Medium performance
4. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, Full HD: Medium performance
5. Radeon HD 5970, E8600 @ 3.33 GHz, HD: Medium performance
6. Radeon HD 4870, E8600 @ 3.33 GHz, HD: Low performance
7. Intel HD4000, i-7 3720 @ 2.6 GHz, HD: Medium performance (without antialiasing; otherwise "low performance")

Performance with shadows disabled:

1. GeForce GTX 780, i-7 4770K @ 3.4 GHz, Full HD: Great performance
2. GeForce GTX 675M, i-7 3720 @ 2.6 GHz, Full HD: Great performance
3. Radeon HD 5730, E8500 @ 3.17 GHz, HD: Good performance
4. Radeon HD 4870, E8600 @ 3.33 GHz, HD: Good performance
5. Radeon HD 6850, E8400 @ 3.0 GHz, HD: Medium perfomance
6. Radeon HD 5730, E8500 @ 3.17 GHz, Full HD: Medium performance
7. Radeon HD 6850, E8400 @ 3.0 GHz, Full HD: Medium perfomance
8. Radeon HD 57x0, i-5 750 @ 2.67 GHz, Full HD: Medium performance
9. Intel HD4000, i-7 3720 @ 2.6 GHz, HD: Good performance (without antialiasing, otherwise "Medium performance" or "Low performance")

I have closed this thread to keep it as a handy reference for everybody. More test results will be added as they are submitted to me by the beta testers.