AMD just announced its highly anticipated Ryzen 3000 processors based on the 7nm node on the very first day of Computex, which proved to be one the most packed keynotes in the history of the event. There were three major announcements, the Zen 2 based Ryzen 3000 processors, the 64-Core Epyc Rome server chip and lastly the RDNA GPU architecture powering the upcoming Radeon RX 5700 graphics cards.
While AMD did have a lot of charts to dis on Intel’s stagnating 14nm Coffee Lake architecture, first-party benchmarks are always taken with a pinch of salt. We’ve already seen a bunch of SiSoft and Geekbench scores, but they don’t exactly give a good measure of real-world performance. The UserBenchmark shared by TUM_APISAK on Twitter, however, is fairly accurate. According to this bench, the hex-core Ryzen 3600 is a staggering 63% faster than Intel’s Core i5-9400 in multi-threaded tasks, 2% in single-threaded in single-threaded and 24% on an average. These tests are based on FP32 based operations, and give a fairly good idea of everyday performance.
The Zen 2 processor is also notably faster than Intel’s 9th Gen i5 chip in gaming (by almost 10%) and given AMD’s pricing history, it’ll also be cheaper than the latter. We are headed towards exciting times. Zen 2, Navi, Ice Lake, Intel’s Xe discrete graphics cards. The dynamics of the PC enthusiast market might as well change in the coming years.