Beware Nvidia: AMD launches new single-slot GPU for data centers – and it looks like an absolute beast
AMD has announced the Radeon PRO V710a new GPU designed for access through Microsoft Azure, its exclusive cloud partner, as part of the company’s family of visual cloud GPUs.
The V710 is available in private preview today, but anyone hoping to use it for a home PC setup will be disappointed, as the V710 has a single-slot design and passive cooling.
AMD has built the V710 on the Navi 32 architecture and there are 54 Compute Units, in addition to 28 GB of VRAM, a memory transfer speed of 448 GB/s and 54 MB L3 AMD Infinity Cache technology. The company says the V710 is ideal for small to medium machine learning inference workloads and small model training via the open-source AMD ROCm software, which aims to compete with Nvidia’s CUDA.
Instance configuration | vCPU | Memory | GPU | GPU memory | Temporary disk size | Maximum bandwidth |
NV28adms_V710_v5 | 28 | 160GB | 1x V710 | 24GB | 1536GB | 20000Mbps |
NV24ads_V710_v5 | 24 | 128GB | 1x V710 | 24GB | 1536GB | 20000Mbps |
NV12ads_V710_v5 | 12 | 64GB | 1/2 V710 | 12GB | 768GB | 10000Mbps |
NV8ads_V710_v5 | 8 | 32GB | 1/3 V710 | 8GB | 512GB | 6600Mbps |
NV4ads_V710_v5 | 4 | 16GB | 1/6 V710 | 4GB | 256GB | 3300Mbps |
The best data center GPU?
“With support for hardware virtualization implemented in accordance with the PCI Express SR-IOV standard, instances based on the Radeon PRO V710 can provide robust isolation between multiple virtual machines running on the same physical GPU and between the host and guest environments,” says AMD . “The efficient RDNA 3 architecture provides excellent performance per watt, enabling a single-slot passively cooled form factor that meets PCIe CEM specifications.”
With Intel gone, AMD has really cornered the market for GPUs at the data center level, and it shows.
Compared to AMD’s RX 7700 XT, the V710 has twice the memory, higher memory bandwidth and a 500MHz lower GPU clock speed. The company says design changes put it at 158 watts, or 35% less power compared to the gaming variant of the GPU.