AMD’s Ryzen 7 family reignited the desktop performance wars last year after nearly seven years of small iterative improvements. The company’s Zen architecture proved more competitive than most expected, and AMD’s top-end desktop cores ruled the multi-threaded roost until Coffee Lake — and arguably past that, given how hard it was to find Coffee Lake in stock during October and November of last year. AMD has already told investors and enthusiasts to expect a new Zen CPU core in the not-too-distant future, but this new revision is a refinement of the existing chip built on GlobalFoundries tuned 14nm process (rebranded as 12nm). The performance jumps from these improvements tend to be smaller, but this is also AMD’s first chance to clean up any low-hanging fruit from Ryzen’s debut and to improve overall performance, which could point the way to larger-than-typical gains.
We don’t know much about how Ryzen+’s clock-for-clock perf will stack up against chips like the Ryzen 7 1800X, but it looks like the revised cores will pack a modest clock speed gain, especially if AMD is more aggressive about holding turbo clocks at high core ratios. If these figures are accurate for the Ryzen 7 2700X, we can make some guesses about a possible Ryzen 7 2800X as well. A 3.9GHz base clock and, say, 4.3GHz boost clock combined with a 4GHz all-core boost would capture an 8 percent clock jump under full load. If AMD has found any small issues to clean up simultaneously — call it 2-3 percent worth of performance — then Ryzen+ could deliver a 7-10 percent gain year-on-year over and above first-generation Ryzen processors.
Is that going to drive an upgrade cycle for existing Ryzen owners? Probably not. But Ryzen+ isn’t intended to reinvent the wheel. Having debuted a brand-new architecture to general acclaim, AMD needs to demonstrate it can iterate on that architecture to continue improving it and that it has picked the right partner to work with on future CPU designs. GlobalFoundries, of course, has its own interest in demonstrating it’s the proper partner for AMD’s efforts as well. The third and final component of this demonstration will be proving that AMD and GlobalFoundries can bring competitive products to market based on GF’s 7nm technology, but that won’t happen until 2019 or 2020. For now, clearing this second hurdle will help demonstrate that AMD will be delivering consistent quality updates over time.
The other reason AMD is likely gunning for higher clocks is so it can open more distance between itself and the Core i7-8700K. When we reviewed the 8700K, we noted that its high single-thread performance and excellent multi-threading perf were enough to make it a stronger overall recommendation than the 1800X. If AMD picks up another 8-10 percent performance, the hypothetical Ryzen 7 2800X should start to open up distance between itself and the Core i7-8700K again.
Finally, these clock speed increases will help reduce the impact of performance deficits that remain between Intel and AMD in specific tests. We don’t know much at this point, but this hypothetical Ryzen 7 2700X ticks all the major boxes — it opens up the throttle, improves performance in the process, and will give AMD a modest competitive boost.
Should Spectre, Meltdown Be the Death Knell for the x86 Standard?
Spectre and Meltdown are serious CPU flaws, but do they warrant throwing out the entire closed-source CPU model?
OnePlus May Have Accidentally Sent Clipboard Data to Chinese Server
The latest beta version of its custom "OxygenOS" Android build was sending user clipboard data to a server in China. Oops.
AMD Releases Updated Risk Guidance on Meltdown, Spectre
AMD has clarified its own weaknesses against Spectre and Meltdown. One week later, things aren't quite as rosy as AMD hoped they'd be.
LG Shifts Strategy, Will No Longer Release Yearly Handset Updates
LG has declared its out of the yearly flagship phone release cycle, preferring to concentrate on retaining form factors for a longer period of time.