AMD's Ryzen 7000 processors are finally out. Zen 4 chips include the Ryzen 5 7600X, Ryzen 7 7700X, Ryzen 9 7900X, and 7950X. Their core count is the same as their predecessors, and most of the work has gone into increasing boost clocks and IPC. AMD claims a performance increase of up to 60% compared to the existing Ryzen 5000 which we will study in detail.
Test bench
- GPU: NVIDIA RTX 3080 Ti FE | AMD Radeon RX 6800 XT.
- Memory: Kingston Fury Beast 16GB x2 @ 6000MT/s.
- Fabric Clock: 3000MT/s.
- Motherboard: MSI MEG X670E ACE | ASUS TUF Z690 Plus D4 | MSI MAG B550 Mortar.
- POWER SUPPLY: ROG Thor 1000W Platinum II.
- CPU cooler: NZXT Kraken X72.
AMD Ryzen 9 7900X vs 5900X: Gaming Benchmarks
We're testing Zen 4 chips on ten AAA titles, including some of the newest with cutting-edge ray-tracing and upscaling technologies. Benchmarks were run at 1080p (Ultra) with FSR 1.0 “Performance preset” enabled as well as RT using CapFrameX. The Ryzen 9 7900X was paired with the MSI MEG X670E ACE, Kingston’s Fury Beast 16GB x2 (6000MT/s | 3000MHz), and the GeForce RTX 3080 Ti. In addition, all ten games were also evaluated with the Radeon RX 6800 XT:
We obtain interesting results here. In rasterized gaming, the Ryzen 9 7900X is barely faster than the 5900X. In contrast, ray-traced titles see gains ranging from 20% in Crysis Remastered to 35% in Hitman 3 compared to the latter. This is due to the increased number of draw calls sent by the GPU in these cases.
The Ryzen 9 7900X's performance is about the same as the 5900X in matrix workloads, but it maintains a healthy 20-25% lead in ray-traced games. In Hitman 3, the 7900X has a massive 40% advantage over its predecessor when paired with the RTX 3080 Ti. The Radeon RX 6800 XT puts all three chips in a tough bottleneck, which is evident in other ray-traced titles.
AMD Ryzen 9 7900X vs 5900X: Content Creation
In content creation workloads like rendering and video editing, the Ryzen 9 7900X shows massive gains over the R9 5900X, despite unchanged thread counts:
Rendering workloads like Blender and Maxon greatly benefit from Zen 4's improvements, gaining nearly 50% performance over its predecessor in most applications. Interestingly, Cinebench benefits the least, a first for a new AMD chip, probably due to the unchanged thread count. V-Ray and POV account for IPC and increased clock counts more efficiently, allowing the 7900X to maintain a massive lead of over 40% over the 5900X.
Compression, Decompression and Encoding
Productivity tasks like compression, decompression, and encoding benefit about equally from core counts and IPC/clock speeds, as shown in the graphs below:
7-zip's LZMA algorithm doesn't like Intel's Gracemont cores at all. The Ryzen 9 5900X is 35% faster than the Core i7-12700K in decompression, and the 7900X, in turn, beats it by more than 40%. In compression, the Zen 4 chip is more than twice as fast as its rivals. Handbrake is more clock/IPC intensive, giving the Ryzen 9 7900X a considerable 50% lead over competing SKUs.
Thermal and power
Despite what early rumors had claimed, the Ryzen 9 7900X, a dual-disk 12-core CPU, performs reasonably well with a 360mm AIO cooler. We tested several options, including the Lian Li Galahad 360 ($200 one piece) without the temperature reaching 70 degrees. CPU power consumption averaged 126W with an upper limit of 136W, even in the most intensive games. The core clocks were also impressive, averaging 5.5-5.6 GHz, even reaching 5.7 GHz in some cases.
Conclusion: Maximum Ray Traced Game Performance
AMD's Ryzen 7000 processors received a lot of criticism after the initial announcement due to the relatively low IPC increase. The launch happened in the blink of an eye and we saw gains of up to 40% in gaming titles and ray-tracked content creation. Essential workloads and productivity workloads also benefit from Zen 4 architecture, perhaps more so than the former.