Benchmarks: Synthetic Performance
Here we see an impressive 30% increase in Cinebench R15 multi-threaded performance over the 5960X, as the 6950X reached 1803pts. As expected, the single-threaded performance wasn't anything impressive at just 147fps. The 6950X was outdone by our 16-core Xeon E5-2670 build by a decent margin.
AnandTech's Ian Cutress created the 3D Particle Movement benchmark quite a few years ago now and it's a great tool for measuring CPU performance. The 6950X scored 1626pts in the multi-threaded test, making it 30% faster than the 5960X and 102% faster than the 6700K. Interestingly, it was again slower than our budget dual-Xeon build, this time by 20%.
The first batch of SPECwpc tests are the product development tests, in which the 6950X was 19% faster than the 5960X while also beating our dual-Xeon build by a slim 5% margin.
The Life Sciences tests favored the 6950X by 21% over the 5960X, while it was also 5% faster than the dual-Xeon system.
The Financial Services test focuses on spreadsheet performance and here the 3950X was a whopping 39% faster than the 5960X and 109% faster than the 6700K. That said, it was 11% slower than the E5-2670 combo.
The general operations benchmark uses applications such as 7-Zip, Octave and Python to measure performance. Here the 6950X matched the 6700K and was 9% faster than the 5960X, while it smoked the 16-core/32-thread budget Xeon build. Clearly this batch of tests doesn't take full advantage of processors with high core counts.