Thus, all 1 to 2 years is to change the time in a computer game thing, but you are addicted to hardware as we need and always have the latest innovation in the system, we recently met a friend at a LAN party, and he marveled at our system. As a brand-new GeForce 7950 GX2 was in the system, we were peppered with questions only Sun One of them was how much CPU power goes out for a 7950 GX2, this is a good and fair question.
Suppose you have a system that is 1 year old and a relatively good X2 3800 + is in the system, with the graphics card you are perhaps not so satisfied, either at the time the money was not enough or the right model is released only a short ones. So one has the question: Should we immediately change the whole system, or is the CPU, a counter question would be, of course, how useful is it to put a much faster processor in the computer?
This has interested and we ourselves have done a few tests, therefore had to serve our X2 3800 +, which we can loosely at 2600 MHz, ie at the level of a X2 5000 + overclock, this difference alone CPU we used for the measurements in games and benchmarks.
From the tests we want to present her to you both synthetic and gaming benchmarks, from here the benchmarks currently offer 3D Mark 2003, 2005 and 2006, of the games ago, we decided for the most demanding 3: FEAR, Quake 4 and X3 Reunion, in this test we want to show the difference of 600 MHz in CPU performance.
From the first benchmark, we can see what makes up much more CPU power, although we have changed nothing on the graphics card, the performance has increased by around 20%, just think about how far you have to overclock a graphics card to get to this value.
In 3DMark 2006 it does not look much different again, we achieved significant growth, which we consider the 3DMark 2006 CPU test also must be having an impact on the score, well you would think, are the benchmarks, but in games, we can not affect so clear, right?
In our first test we have used FEAR games, since we have used the same measurements with high quality, which means 1280x1024 and FSAAx4 and Afx16, the difference amounted to about 14 frames on average, more faster for the CPU. Even when 65 frames with 3800 + perfectly adequate, were almost 80 frames are on average much more, as you can see, a graphically complex game is throttled by the CPU when not enough MHz.