Radeon Pro v340

AMD Radeon™ Pro V340

Delivering smooth GPU experience from the cloud to virtually any device, anywhere.

New AMD Radeon Pro V340 is a monster card that delivers virtual GPU to VDI deployments. GPU space is becoming exciting again.

– Steve McDowell, Moor Insights & Strategy


The AMD Radeon™ Pro V340 datacenter graphics card delivers an impressively smooth GPU experience from the cloud to virtually any device, anywhere.  With MxGPU technology at its core, this hardware-based virtualized graphics solution provides high levels of predictable performance, enhanced security and support for up to 32 VMs per card.  IT managers will also like the fact that this MxGPU solution is easy to set up and manage. Unlike other GPUs, the V340 delivers GPU virtualization to end users without end user licenses, providing enterprises with a lower cost per user.

AMD MxGPU Technology in Action

AMD showcasing our latest and greatest virtualization solutions at VMworld 2018

Radeon Pro v340 @ VMWorld 2018
Radeon Pro v340 @ VMWorld 2018
Radeon Pro v340 @ VMWorld 2018

Key Features

MxGPU Hardware Virtualization Technology

Up to 32 Virtual Machines (VMs) per card

Two Virtualized Encode Engines to compress independent video streams in H.264 or H.265 formats

Remote Management tools to monitor a range of static and dynamic GPU information

Two GPUs based on the advanced “VEGA” architecture

56 x2 CU COMPUTE UNITS to accelerate demanding workloads1.

State-of-the-art memory technology: 32GB of HBM2 MEMORY

Enhanced Security Engine to enable hardware isolated VMs

AMD MxGPU Technology

Pure. Datacenter. Graphics.

Discover the benefits of the world’s first hardware-based virtualized graphics solution

Radeon™ Pro V340

Compute Units:
Stream Processors:
Memory Size:
32 GB
Memory Type (GPU):
Memory Bandwidth:
484 GB/s

Radeon™ Pro: Stay up-to-date with the latest from AMD

Radeon™ Pro V340 Graphics Press Release

See what Michael DeNeffe has to say about the Radeon™ Pro V340 Graphics

  1. Discrete AMD Radeon™ and FirePro™ GPUs based on the Graphics Core Next architecture consist of multiple discrete execution engines known as a Compute Unit (“CU”). Each CU contains 64 shaders (“Stream Processors”) working together.  GD-78