
As with all unlocked K SKUs, the -7350K doesn't include a heat sink. Regardless, the prospect of an unlocked Core i3 is exciting, and it shows Intel listening to the post-Pentium G3258 feedback. The increased capabilities of HD 630 graphics make the on-die solution usable, but a stingy 4MB 元 cache is a bummer. A 4.2 GHz base clock rate is encouraging, and the lack of Turbo Boost shouldn't bother tuners much. Obviously that makes it an interesting option for overclocking on the cheap. Intel's Core i3-7350K marks the company's first unlocked dual-core/quad-threaded model with a 60W TDP. The Core i5-7600K is a great deal for the value-conscious, but an unlocked Core i3 adds a new wrinkle to the locked-i5 purchasing decision. Thankfully, Intel left its pricing structure alone, and the Core i7-7700K retails for $339 while the versatile i5-7600K weighs in at $242. Turbo Boost binning is also impressive, ranging from a single-core 4.2 GHz to 4.0 GHz on all four cores. The four-core/four-thread Core i5-7600K stands out as an impressive value with its 3.8 GHz base clock, which is up 300 MHz faster compared to its predecessor. The -7700K's base clock rate is 4.2 GHz, an increase of 200 MHz over the previous-gen -6700K, and all four Turbo Boost levels jump by 300-400 MHz. The MSI motherboard we used as a test platform also features an Enhanced Turbo feature that increases the four-core frequency to 4.5 GHz. We also saw the chip sustain 4.4 GHz across all four cores during our stress tests. The four-core/eight-thread Core i7-7700K catches our eye with a top Turbo Boost frequency of 4.5GHz. We expected Intel to add Turbo Boost 3.0 support, which identifies single-threaded workloads and pins them to the fastest core, but it's apparently keeping that feature exclusive to Broadwell-E. We're dealing with the same dual-channel memory controller from Skylake, though Intel bumps official DDR4 support up to 2400 MT/s from 2133 MT/s. Intel also moved to the slightly-enhanced HD Graphics 630. Not surprisingly, we find the same allotment of IA cores, TDP ratings, and cache hierarchies. Kaby Lake processors drop into the LGA 1151 interface, so you can use them with existing Z170-based motherboards (after a requisite BIOS update) or the new Z270 chipset. As we learned during Intel’s mobile Kaby Lake launch, most of the changes come in the form of an improved media engine, a nimbler SpeedShift, and the welcome addition of higher frequencies. So we aren't the lucky recipients of any IPC enhancements or other new technologies. However, the previous-gen architecture rides back into the picture atop the process improvements. These CPUs are borne from targeted refinements to the existing node, resulting in faster transistors that enable 300-400 MHz-higher clock rates than Skylake could achieve. Instead of plowing forward with a 10nm shrink, Kaby Lake falls into the optimize phase and comes bearing Intel’s tuned 14nm+ process. Intel is now adhering to its new Process-Architecture-Optimize cadence, or tick-tock-tock, if you will. It marks a departure from the heady days of its Moore’s Law-fueled tick-tock cadence, which normally finds the company presenting a new microarchitecture followed by a die shrink. Today's desktop-oriented Kaby Lake launch is different, though. And in the face of an impending AMD Ryzen roll-out, Intel really wants power users to get excited about a refreshed processor portfolio. The company’s insistence that it isn’t abandoning enthusiasts is genuine.
