Sign in with
Sign up | Sign in

Mass Effect

Radeon HD 4870 X2: Four Cards Compared
By

With the overclocked quad-core CPU (QX@3.67), Nvidia gets a strong competitor. While AMD needs no less than the Radeon HD 4870 in CrossFire (CF) or the dual-chip Radeon HD 4870 X2 in order to catch up to the GeForce GTX 280 or even beat it, a GeForce GTX 260 is easy prey. The UT3 engine is perfect for showing 3D power (GPU scaling), and the Radeon HD 4870 X2 CrossFire (4CF) can make use all four of its graphics chips to constantly lay down very playable frame rates. A faster CPU will raise the frame rate in all resolutions, although it’s hard to say where the limit is.

The fastest graphics cards in SLI and CrossFire modes are limited by the CPU in all resolutions without anti-aliasing. With anti-aliasing, the graphics power makes a difference. SLI with GeForce GTX 260 and GTX 280 cards only pays off in higher resolutions, while the single card wins in lower resolutions. When combined via SLI, both GTX cards are limited by the CPU—you can see that from the fact that there are hardly any differences between the chips, and the weaker GTX 260 can easily follow the GTX 280.

Almost all single-card solutions are taken to their limits with Mass Effect. The Radeon HD 4850 shows minimal or no increase whatsoever from the overclocked CPU. GeForce GTX 260 and GTX 280 cards hardly benefit from the fast quad-core CPU above 1680x1050 with anti-aliasing. The Radeon HD 4870 shows only low gains in frame rate above 1920x1200 without anti-aliasing.

Driver Comparison: Catalyst 8.9 Versus Catalyst 8.10 w/ Radeon HD 4870 X2

You can see a slight increase in Mass Effect, but most of the results are within the normal measurement variations.

Mass Effect in fps 1280
0xAA
1680
0xAA
1920
0xAA
1280
4xAA
1680
4xAA
1920
4xAA
1280
8xAA
1680
8xAA
1920
8xAA
Overall
HD 4870 X2 8.10 QX@3.67 131.2 129.7 125.0 116.2 87.6 70.6 114.1 85.7 69.0 929.1
HD 4870 X2 8.9 QX@3.67 129.6 128.3 126.0 113.7 85.9 68.6 112.5 83.9 66.7 915.2
Performance difference in percent 1.2 1.1 -0.8 2.2 2.0 2.9 1.4 2.1 3.4 1.5

React To This Article