It’s difficult to draw conclusions about hardware still shrouded in mystery. Although there’s plenty we know about Skylake based on Intel’s official disclosures, information leaked elsewhere and the diagnostic tools we use in the lab, this launch just doesn’t feel complete without the background we’d expect to accompany a new architecture. Fortunately, we have the fundamentals: we know how Core i7-6700K and Core i5-6600K perform, we know how much they cost, we have a good sense of their strengths and we’ve seen the platforms they’ll drop into.
Arguably, motherboards are the most exciting part of the story (when was the last time you heard that?). The changes introduced by Intel’s Z170 chipset pave the way for an era of PCI Express-based storage, blasting through the throughput and latency limitations imposed by SATA 6Gb/s and AHCI. Of course, it’s nice to see greater proliferation of USB 3.1 controllers, DDR4 support and granular BCLK overclocking as well.
The Core i7-6700K has a lower Turbo Boost ceiling than Core i7-4790K. But its superior IPC translates to better performance in most single- and multi-threaded workloads. Under the heaviest tasks it typically only trails six- and eight-core Haswell-E parts. Of course, the benefit of Intel’s mainstream platforms is their more modern feature sets. So do you want all of the features I just listed off, or do you need lots of processor-based PCIe and more cores?
Intel will go into more depth on Skylake during IDF later this month, and I don’t see any reason to rush into a purchase before then (provided processors surface for sale right off the bat, that is). Should everything we learn support the data we generated today, then I think it’s safe to say Skylake will become the first architecture to really get enthusiasts excited since Sandy Bridge—and not even entirely because of the processors themselves.