In November 2008, the first generation of processors Nehalem, and at the end of March 2009 the company began to move Intel Nehalem architecture in the server segment. The new Xeon chips are designed for single or dual systems. Recall that in the Nehalem architecture has been implemented a number of important innovations, including an integrated memory controller, gipertreyding with support for up to 16 virtual cores and Technology Turbo Boost. Go into more detailed technical specifications and technology, we will not, and will test new items. It will do this using tools that are understood by most users. Let`s say, will measure the productivity of server chips with the help of applications for the desktop segment. My first acquaintance with the Intel Xeon 5560 occurred during the launch Xeon 5500 series in Ukraine. The processor runs at 2, 8 GHz and has 8 MB of cache memory of the third level. With the Turbo Boost ability to dynamically increase the operating frequency of individual cores to 3, 2 GHz. As expected, after the official part of those present were invited to meet with the ready-made solutions based on server processors Nehalem partner Intel. Given the limited time we have been able to make a few screenshots and a test run wPrime 32M. The result of this application has been impressive. Rating Site HWbot.org, he hit the TOP 20. After some time we have been able to test the server-based two-processor Xeon 5560. The system was assembled at the base of the motherboard Intel S5520HC, two slats DDR3 memory volume to 1 GB GeForce 9800 graphics card, and GT. As the operating system was chosen as a familiar Windows Vista SP1 x86. So, in order to detect the work of Turbo Boost technology was launched on a test SuperPi 1M. Output 13, 057 with anything interesting is not represented, but the frequency of 3, 2 GHz instead of standard 2, 8 GHz, have fun. The next test will allow us to understand how the growth will be a system of technology gipertreyding. To this end, it was running wPrime 32M with a miscalculation in the 8 and 16 flows. With eight errors in the flow of the system handled over 7, 846 s, and at sixteen maquila computation for only 4, 54 pp. In other words productivity increased by 70%. During testing in wPrime 1024M managed to take 17 place in the world ranking with the result in 134 pp. Try this test on a home PC. Test package CineBench R10 perfectly optimized for multicore processors. When turned off in BIOS Hyper-Threading Technology has received the result equal to 20065 points. When the result of this technology has increased to 22, 718 points. The difference is not great, but 13% increase, roughly speaking, on a flat spot, a good tool. In CineBench 9. 5 is obtained about the same difference of activation Technology Hyper-Threading. Eight streams - 2855 points, 16 flows - 3102 points. We filmed a video that shows the rendering of images that display the bike in 16 streams. Since the processor tests of the applications 2006, and 3DMark Vantage is not so simple. The point is that the application gave the error. The first test runs and the second error. As it turned out, the reason for hiding in the Technology Hyper-Threading. Disable it, managed to pass all tests. Why exactly do not know. Of course, testing the server with the help of applications for the desktop segment is not entirely correct. The first processor test 3DMark Vantage without Hyper-Threading Technology was received 3857 points, but when it activates 4853 points. Growth was 25%. But as many rings prepared 3DMark Vantage to the second test with 16 threads. Turbo Boost Technology and Hyper-Threading on. Performance is impressive. We can only dismiss these processors. And look at what they can do.