Unfortunately, HandBrake wasn’t able to fully utilize the 16-core setup; CPU utilization only hovered around the 30% mark, which means that it was effectivly using around 10 threads. Still, we see a reasonable gain in performance and the Xeon E5-2670s are still able to outperform the 5960X.
Since one copy of HandBrake couldn’t make full use of the dual Xeon E5-2670 processors, we decided to run three simultaneous instances to see how the 32-thread system handled the load. This time around we see a nice 50% performance gain going from a single E5-2670 processor to two. Whereas the dual Xeons were just 6% faster than the 5960X when running one instance, they are now 25% faster running three instances. As you can see, the more work you can simultaneously throw at this setup, the better it does.
The dual Xeon E5-2670 processors were put to good use in the Hybrid x265 encoding test as they allowed for a 72% performance boost over a single processor. However, despite the massive performance boost, they were just 15% faster than the 5960X.
Unfortunately, the x264 HD Benchmark wasn’t able to properly utilize the 16-core setup, and, as a result, the dual Xeon E5-2670 processors were slower than the 5960X in the second pass test. That said, they were faster in the first pass test by a 14% margin.
