Core i9-9900KS Web Browser and Encoding Performance
We include stock testing results of the Core i9-9900KS with both our custom loop (CL) and Corsair H115i cooler (H115i).
|DDR4-2666||Core i9-9900KS CL, Core i9-9900KS H115i, Core i9-9900K, Ryzen 7 3700X, Ryzen 9 3900X, 9900K @ 5.0 DDR4-2666|
|DDR4-3600||Core i9-9900KS @ 5.2, Core i9-9900K @ 5.0, Ryzen 7 3700X, Ryzen 9 3900X|
Browsers tend to be impacted more by the recent security mitigations than other types of applications, so Intel has taken a haircut in these benchmarks of fully-patched systems.
The SVT-AV1 encoder is an Intel- and Netflix-designed software video encoder that became available earlier this year. This new encoder is more scalable than other encoders, thus offering faster performance paired with efficient compression. While it may seem counterintuitive to use an Intel-designed encoder for testing AMD processors, consider that most encoders are inherently reliant upon per-core performance, which is a strength of Intel, while SVT-AV1 exposes the power of threading, a strength of Ryzen. At stock settings, the Ryzen 9 3900X dominates the test pool, but the auto-overclocking PBO feature does little to improve performance. We see expected gains from the -9900KS after overclocking, and the Core i9-9900K with tuned memory shows that memory data transfer rates have an impact on this test.
Our LAME and FLAC tests, like many encoders, rely heavily upon per-core performance. That means Intel's frequency advantage confers a big lead.
Intel processors traditionally leverage high frequencies to dominate the HandBrake x265 test, which relies heavily on AVX instructions, and the x264 test. But Intel's higher clock speed isn't too much of an advantage in these tests when the competition has twice the number of threads.