Prev3 of 5Next

Rendered From Nothing

Made by a company by the name of MAXON, CINEBENCH is rendering benchmark that tests your CPU and GPU for performance in 3D content creation. In the care of mutli-core or hyper-threaded CPUs, CINEBENCH will test both single and multiple core performance.

In both these tests the Phenom II X4 965 posted some more levelled scores. The single CPU score was about 6% faster and the multiple core score was about 5.5% faster. The missing half percent when compared to the increase in processor speed could be due to some playing about with the GPU. Still for a CPU using the same core but only 6% faster, these scores are what was expected.

3D rendering is all fine and dandy, but more users are rendering their own live video these days. Unfortunately it’s hard to find a rendering tool or program that takes full advantage of multi-core CPUs. For this reason we turn to Nero Recode 3, which can take any unprotected DVD and convert it into any number of video formats. For our testing, we chose the original Stargate movie for rendering. It may be old, but the many bright scenes and plentiful action give the rendering tools a good workout.

Once testing was completed (1.5 hours later since video rendering still takes time), we once again found that the 965 had pulled ahead of the 955 by a percentage close to its increase in clock speed. In this case the 965 was 5.5% faster then the 955.

Playing Away The Day

Now I know I said I wasn’t going to do any game benchmarks, and technically I’m still not. At least I’m not doing any real world game tests, which I’ll elaborate on in a moment. I still wanted to run through 3DMark Vantage and see if there was any performance different at all between the two CPUs, or if it’s really all on the graphics card. Turn’s out it’s all on the graphics card.

With the exception of the High performance test, there was less then 1% different between the 955 and 965. The High test seems to be an anomaly of averages. Gaming only seems to benefit when there’s new optimizations in a CPU, or a new core altogether. A 200MHz speed increase isn’t going to do it.

As for other games, the same story was repeated. In the two titles we use for CPU and mainboard testing, there was a statistically insignificant difference between the two CPUs.

Is It Hot In Here?

Since we plan on overclocking this CPU, I figured it was  good idea to see what kind of thermal performance it has. After all we are working with a high TDP on the Phenom II 965. Turns out that at idle the 955 and 965 are near identical. The 965 is always within a half degree Celsius of the 955. Using OCCT to place 100% load on all four cores I found that the 965 does get a little hotter than the 955, if only a little. We can see a 1.5°C temperature increase in the 965 at full load. So in theory this CPU should have lot’s of room to overclock. I think we should find out.

Prev3 of 5Next

Share This With The World!