VR-Zone said:So, firstly, there is the large 12-core die with 30 MB L3, where three 4-core columns share two sections of L3, 20 + 10 MB, across three ring buses. In our first review, the Xeon E5-2697v2 2.7 GHz is the top bin part representing this die. You’d think this monster drinks power like a sweaty runner after a marathon? Well, guess what, measured at the power supply, the whole dual-CPU 24-core system with 128 GB DDR3-1866 RAM and enterprise SSD, plus a high end AMD7990 GPU, Consumed less power than the brand new 4-dualcore AMD-9590 single socket system with 8x less memory and similar graphics, for a roughly ~4x system level performance gain!
How would a single XeonE5 2697 v2 stack up against the likes of a 3930k in terms of raw computational power? I was potentially thinking of convincing my PI to purchase either the 12 core, or the 10 core version to speed up our mathematical computations in quantum chemistry. We currently like the X79 platform as we need lots of memory and a 3930k is fast, however, these seem like a good consideration. Would it be at least a 60% improvement? (I know a 2x improvement is unlikely) However, this seems like more than just an incremental improvement by intel to me.
Wouldn't you get better performance with a cluster of cheaper machines? With the money of that 12core cpu plus the platform you could buy many more cores.
A "cheap" cluster quickly get rather expensive hardware, complexity, scaling and TCO wise. Plus you lack RSA feastures. But again, the real killer is simply software licensing.