I read that xeon 12cores 2697 2.7GHz will score around 1500points in cinebench and a new 5960x 8core 4.4GHz oc will give 1700cb points.
Does this mean this 8core will render faster than the 12 core? Might be a dumb question, but i read cb points is not necessarily true to actual rendertime.
First of all, Cinebench is an extremely accurate tool to compare CPUs speed, and thats why it’s one of the most used industry standard.
The 5960 will get 1700 points only if heavily overclocked(and you can get this only if you are lucky enough to get a good one). Anyway, even if 1500/1600points seems a more stable/reasonable result, if you consider that the 2697 costs 2.5k and the 5960 costs just 1k, there’s no match(unless you really need ECC for specific tasks, or dual CPUs configuration).
Also consider that the 5960 will be much faster in most tasks due to better single thread performance.
You’re not really comparing apples to apples here. The Xeon E5-2697 v2 is meant to be used in dual processor configurations so you’d have 24 processor cores (around 3,100 score on Cinebench R15). Out of the box the i7-5960X will do about 1,300 in Cinebench R15.
In terms of ray tracing rendering performance the dual processor Xeon options are way faster but also a lot more expensive. Cinebench is a good indication of the ray tracing performance to expect from a processor but it doesn’t take into account all the factors of real world production like large data sets and possible bottlenecks like network storage. So while one option might be 50% faster or 100% faster in the benchmark doesn’t mean it’ll actually be 50% faster or 100% faster in real world production.