|
Palit - ATI Radeon HD 4870
Testing Rig
Processor |
Intel - Q6600 2.4 GHz @ to 3.51 GHz |
RAM |
|
Motherboard |
Gigabyte - GA EP45 DQ6 |
Graphics Card |
Palit - ATI Radeon 4870 |
Cooling |
Thermaltake - Bigwater SE |
Case |
|
PSU |
3D Mark
Single Card |
CrossFire |
![]() |
![]() |
3DMark results only point to one thing that may be obvious to you: 2 cards are better than 1. The score of CrossFire set-up reached nearly twice that of a single card in overall and graphics scores as well as frames per second. Taking a closer look at the pictures above, one can notice that CrossFire configuration played the video demos almost twice the rate of the single card. This is great news for those of you who are planning on buying 2 of these bad boys for you PC as it will definitely have a positive impact in your computer's gaming performance as you will see later in the review.
PCMark
Non - Overclocked Scores |
|
Single Card |
CrossFire |
![]() |
![]() |
The difference between CrossFire and a Palit's Radeon 4870 set-up was not as grand as I hoped to for. In fact, I admit I was a little disappointed that the CrossFire set-up performed a mere ~100 points better than the single card set up. Baffled by this difference, I decided to investigate into this matter some more with more testing. The scores that you see posted above came from my machine at default settings, not overclocked. The overclocked scores surprised me even more! This time the single card scores were higher than the CrossFire set-up. The scores were obtained of course after tweaking the bios for over 2 hours and having the benchmark fail over and over again. Nevertheless I was able to get the system stable enough to run the tests, and here are the results:
Overclocked Scores |
|
Single Card |
CrossFire |
![]() |
![]() |
As you can clearly see the scores do not quite match up. This could be because of several reasons. First and foremost is the power supply. If your power supply is not up to par with the full consumption of the graphics cards then the benchmark will fail. Unfortunately for me that was not the case since I used an 850 watt power supply. This brought me to the second and more probable reason. The motherboard that I am using is the GA EP45 DQ6, has a 16x PCI port for the first graphics card and 8x PCI port for another graphics card. This means that whenever I am using the graphics cards in CrossFire configuration, I am bandwidth-limited to only 8x instead of the 16x (max capability of the graphics card). There is a possibility that the bandwidth limitation might have a negative effect on the overall stability of the system thereby reducing the score.
CINEBENCH R10
Single Card |
CrossFire |
![]() |
![]() |
During this test 4870s in CrossFire configuration exhibited similar results as those in the PCMark. The scores from the CrossFire were lower than those of the single card. This is most likely connected to the stability issues with the motherboard or perhaps the way the program was written.
REALTIME PRICING