As TechSpot notes, the 2700X’s hits 9387 when overclocked to 5GHz, while the 8700K tops out at 8935. 3DMark isn’t normally an application we’d turn to for CPU performance figures, but Time Spy is clearly more responsive to improved CPU performance than conventional tests like Fire Strike. Intel is also expected to pair this chip with the new Z390 chipset, which will improve on the Z370 in two respects: The Z390 adds support for six USB 3.1 Gen 2 ports, and it adds an integrated 802.11ac Wi-Fi modem. Previously, high-end boards from companies like Asus that offered this kind of support were using third-party products to do it or integrating a separate Intel Wi-Fi module onboard.
There are a few takeaways here. First, it’s not surprising that Intel’s Core i9-9900K would outperform the Ryzen 7 2700X — in fact, it’s expected. Intel has labs of equipment dedicated to this kind of testing and they aren’t going to launch a $500 chip — if that’s what the Core i9-9900K costs — if they can’t back up the performance of it at some level. The question is, will AMD respond with a new top-end processor of some sort, or will they hold their own response until Ryzen 2 debuts on 7nm?
Intel does, after all, still have a card it could play in all this. With Coffee Lake, Intel chose to bump core counts across its entire mainstream product lineup. While it seems unlikely, Intel could theoretically bump the i3 and i5 families by adding HT to both. Whether the company would practically take this step is an open question; such a move would put the Core i3 on par with a Kaby Lake-era Core i7. Intel could compensate for the shift by dropping clock speeds to create space between product SKUs going forward.
But for now, this seems like a premature move for Intel to make. Our guess is that the new 9th Generation CPUs will keep the existing Core i3 / i5 stack and adopt a new strategy for the Core i7 and Core i9. That way, if AMD’s Ryzen 2 is a major leap forward, Intel can still respond by pushing core counts higher on the i3 and i5 families at a later date. It’s not just a question of when Intel’s 10nm will be ready. While that topic has dominated the discussion in recent months, some of you may recall that Intel originally planned to keep using 14nm chips on the desktop until 10nm+ was ready for prime time. We don’t know how the 10nm delay will impact that timeline or how Intel will upgrade future product families. It may also depend on whether AMD is willing to push core counts even higher at that $500 price point to compete.
Enthusiasts, rejoice! The CPU market looks likely to stay fairly interesting through the back half of 2019, at least.