AMD officially announces the Radeon PRO VII workstation graphics card, promising to become a tough competitor to NVIDIA GV100 and GP100 models and the Quadro series, not only in performance but also in price.
AMD Radeon PRO VII – 16GB HBM2, PCIe 4.0, 1TB/s bandwidth, 8K and more
AMD Radeon PRO VII is a graphics card designed for content creation. It is optimized for 8K and highly scalable, so we can combine several of these cards to improve performance and memory.
The workstation graphics card offers 16GB of HBM2 memory, PCIe 4.0 support, 6 DisplayPort connections, and bandwidth of over 1TB/s. AMD promises single precision performance (FP32) of 13.1 TFLOPs and 6.5 TFLOPs (FP64) with double precision.
AMD compares it to the NVIDIA Quadro RTX 5000 and RTX 4000 and highlights the advantages not only in bandwidth but also in performance and power consumption with support for 6 DisplayPort displays.
When two Radeon PRO VII are combined, we can have 32 GB of memory available, this amount is comparable to the GV100 and GP100, but costs much less. Compared to the GV100, the cost savings are about $5200 and the performance is almost twice as high on the FP64.
With the introduction of this new Radeon PRO Radeon VII graphics card, ProRender 2.0 will also be introduced. Unreal Engine, SideFX, Autodesk Maya, and Blender confirm their support for the new rendering algorithm, which promises to improve performance and workflow with Radeon PRO. There is also support for Infinity Fabric, which scales performance better when two or more graphics cards are combined in parallel.
The Radeon PRO VII will have a consumption of 250W TDP and use a 6+8 pin power supply.
Its official price is 1899 USD, which is lower than that of its direct competitor, the NVIDIA Quadro RTX 5000 in this segment. We will keep you up to date.