For this review, we decided on a fresh start, which meant replacing some of the older components in our system configuration. Since the Intel processors are using the Socket LGA1156 package, we went with a MSI P55-GD85 motherboard that has the latest features, like SATA 6Gbps and USB 3.0. The memory and graphics card remained the same, since they are still adequate for our needs. However, our old Seagate Barracuda 7200.10 hard drive has been replaced by a Western Digital Caviar Black 1TB SATA 6Gb/s drive. We also decided that a switch to a modern 64-bit operating system was long overdue, hence the choice of Windows 7 Ultimate (64-bit). Similar tweaks were made to our AMD test system, like the HDD and the OS, as the ASUS Crosshair IV Formula (AMD 890FX) is recent enough to stay.
For our comparisons, we decided that these K-series processors should be compared directly to their non-K versions, so the Core i7-870 is tested with the i7-875K. However, there's no need to duplicate the same for the Core i5-655K, so we threw in processors that are in the similar price range, which include the US$200 Core i5-661 and the Core i5-750. Next, AMD has quite a few choices in price brackets close to the Core i7-875K and i5-655K, so we included them too, like the US$296 Phenom II X6 1090T and the US$200 Phenom II X6 1055 and the US$180 Phenom II X4 965. To recap the prices of the processors we're comparing here:
What's immediately obvious here is that the Core i7-870 looks extremely overpriced compared to the new i7-875K. Given the similarities between the two (except for the virtualization and enterprise bits), there's hardly a reason to get the 870 now when the 875K comes with similar performance and an unlocked multiplier. However, the Core i5-655K looks less of a bargain, with a price premium over the higher clocked Core i5-661, which has a higher clock speed (but no unlocked multiplier). And of course, the AMD contingent looks very price competitive too. It will be interesting to see how the benchmarks turn out.
With a revamped system configuration, we have also updated some of our benchmarks to reflect this, adding in 64-bit versions when available. We also replaced our aging XMpeg benchmark with Handbrake to test H.264 encoding. The following benchmarks were used in this review: