1. Our test method is described here.
2. All data types are reduced to one relative integer score: performance of a reviewed processor compared to performance of Intel Core 2 Quad Q6600, which is considered 100 points.
3. The detailed results are available in this Excel spreadsheet. The review contains only summaries for each benchmark class.

3D modeling and rendering

Here is a comment on this group of tests from the Xeon X5560 review: deviations from our assumptions are quite significant: if not for the optimizations of the last months, this result should have been demonstrated by at least X5570, or even the hypothetical Xeon 3.06GHz. As we can see, it's not about optimizations in the first place, these tests depend much on memory performance. To be more exact, in tests that mostly deal with computations (rendering in particular), processors rank according to their clock rates. In the other tests, the difference between the Core i7-950 and Xeon X5560 is infinitesimal. The same concerns the total score -- almost a 10% difference in clock rates helped the first processor gain only a 2% advantage. Well, it's another proof that latencies play a significant role for processors with the integrated memory controller. It was not without a reason that AMD did not want to support DDR2 before it mastered 667MHz. The same happens with Intel now -- DDR3-1333 is sometimes much faster than DDR3-1066. Desktop Nehalems are artificially restricted to the latter only because the company does not want to let people overclock them too much, clearing the road for the Core i5.

CAD/CAM

This tendency is even more pronounced here. Especially owing to UGX NX -- if we take a look at the detailed results, we'll see that the 950 is outperformed by the X5560 in one of two subtests. And in the second test it's faster only by a measurement error value. However, performance comparison in the 920-950-965 series shows positive prospects for the new device: everything is proportional to the core clock rate, so it's only insignificantly slower than the former top processor (950).

Compiling

Test results depend directly on the clock rate only. Conclusion? Core i7-950 is an excellent processor for its price. It will impact the family of single-socket Xeon 3500 processors, if the company does not do the same castling here.

Professional photo processing

Photoshop depends on all parameters, but the CPU clock rate is still a tad more important here than memory frequency. Core i7-950 and Xeon X5560 demonstrate identical results in some tests, but the overall victory goes to the higher-clocked processor.

Scientific applications

Core i7 is only insignificantly faster than Xeon X5560 again.

Web server

CPU clock rate is very important here, although memory frequencies also have certain influence on results.

Total professional score

As we can see, the new top 'regular' processor from the Core i7 series does not outperform much the lower-clocked processors from the professional series in professional applications. But it's only a little slower than the old top processor from its family. Remember that it will have the same price tag as Core i7 940, and extreme models will be almost twice as expensive. Before the rollout of 'true' desktop processors with the Nehalem core, the company can make only such presents to its customers. Nevertheless, it's a very good present -- a tad higher performance for the same money. This processor has almost reached a performance level that had been the highest for half a year. These presents are only for new customers, of course. But that's how all price cuts and new models work.

Write a comment below. No registration needed!