AMD Radeon VII Review: Performance Benchmarks With 7nm Vega

How We Configured Our Test Systems: We tested the graphics cards represented in this article on a Gigabyte Aorus X299 Gaming 7 Pro motherboard powered by an Intel Core i9-9980XE 16-core processor and 32GB of G.SKILL DDR4 RAM clocked at 2,666MHz. The first thing we did when configuring the test system was enter the UEFI and set all values to their "high performance" default settings and disabled any integrated peripherals that wouldn't be put to use. The memory's clock and timings were manually dialed in to ensure optimal memory performance at the processor's maximum supported speed of 2,666MHz (without overclocking), and the solid state drive was then formatted and Windows 10 Professional x64 was installed and fully updated. When the Windows installation was complete, we installed all of the drivers, games, and benchmark tools necessary to complete our tests.

We should note that the AMD Radeon RX Vega card was tested in its default "Balanced" power mode throughout. Power Saver (slower) and Turbo (faster) power modes are also available with Vega, which would affect performance, noise output, and peak power consumption.

HotHardware's Test System Intel Core i9 Powered

Hardware Used: Intel Core i9-9980XE (3.0 - 4.3GHz, 18-Core) Gigabyte X299 Gaming Pro 7 (Intel X299 Chipset) Radeon VII Radeon RX Vega 64 GeForce GTX 1080 GeForce GTX 1080 GeForce GTX 1080 Ti NVIDIA Titan Xp GeForce RTX 2060 GeForce RTX 2070 GeForce RTX 2080 32GB G.SKILL DDR4-2666 Samsung SSD 860 Pro Integrated Audio & Network

Relevant Software:  Windows 10 Pro x64 NVIDIA Drivers: v417.71 AMD Drivers: Crimson v19.1.1 / PRESS Benchmarks Used: SiSoft SANDRA GPGPU LuxMark v3.1 Unigine Superposition VRMark 3DMark "Fire Strike" 3DMark "Time Spy" Middle-Earth: Shadow Of War Shadow Of The Tomb Raider F1 2018 Strange Brigade Final Fantasy XV FarCry 5

SiSoft SANDRA 2018 Video Shader Performance Tests

First up, we have the Video Shader Compute and GPGPU Shader Processing benchmarks built into SiSoft SANDRA 2018. These tests performs a series of floating point operations on the GPU and reports the average throughput of the calculations. The Shader Compute test uses DirectX, while CUDA or OpenCL is employed for the NVIDIA GeForce or Radeon cards, respectively, in the GPGPU Shader Processing tests...

The Radeon VII lands right about in the middle of the pack in the single-precision test here, but absolutely blows past every other card in the double-precision test, nearly tripling the performance of its nearest competitor, the Radeon RX Vega 64. Thanks to that massive lead in the double-precision test, the aggregate score also shows Radeon VII with a huge lead over all of the other cards. 

The Shader Processing benchmarks tells a very similar story. The Radeon VII's single-precision performance is in-line with the GeForce RTX 2080, but it trails the Titan Xp or GeForce RTX 2080 Ti. In the double-precision test, however, the Radeon VII blows past everything else by a wide margin, which again pushes its aggregate score into the stratosphere.