AMD announces the Radeon Pro V340 GPU with 32 GB HBM2 VRAM for VDI applications

AMD Radeon Pro V340 dual GPU. (Source: Tom's Hardware)
AMD Radeon Pro V340 dual GPU. (Source: Tom's Hardware)
At the ongoing VMworld 2018 conference in Las Vegas, AMD announced the Radeon Pro V340 dual-GPU for Virtual Desktop Infrastructure (VDI) applications. The Radeon Pro V340 can power demanding datacenter workloads such as CAD, rendering, and Desktop-as-a-Service (DaaS) applications. The V340 can support up to 32 VMs, each with its own independent memory access for extra security along with support for independently compressing H.264 and H.265 video streams.

A few months ago, we reported about AMD's plans for the high-end datacenter market with the Radeon Pro V340. Now, AMD has made the V340 official at the VMworld conference yesterday in Las Vegas. The V340 is not your typical gaming or even a workstation GPU but one that is aimed at Virtual Desktop Infrastructure (VDI) environments. A VDI provides clients including desktops, thin clients, and mobile devices with a personalized desktop that mimics a local installation except that it resides in a datacenter.

The Radeon Pro V340 is based on the Vega architecture and is a dual GPU solution made possible by AMD's MxGPU technology. It is still not the 7nm Vega that AMD showed off at Computex 2018, though. MxGPU works on SR-IOV that allows up to 16 virtualized users per physical GPU. Therefore, the dual Vega GPUs in the V340 can support up to 32 concurrent virtualized users. The V340 sports 32 GB HBM2 ECC memory with each user getting 1 GB memory for remotely running demanding workloads such as CAD simulations, rendering, and Desktop-as-a-Service (DaaS).

AMD's MxGPU enables scalability without having to incur recurring hardware licensing fees and a datacenter can leverage as many such V340s as needed. AMD says its user density is 33% more than the competing Tesla P40 offering from NVIDIA that supports up to 24 concurrent users.

Since the data doesn't reside on the client, a VDI setup is inherently secure and the V340 further augments this with a built-in security processor that provides secure boot and data encryption. Each connected client is isolated from the others at the hardware level. The GPU also can independently compress both H.264 and H.265 video for all the shared users. The exact GPU specifications, however, were not yet revealed. 

Pricing details were not revealed but AMD says that the Radeon Pro V340 can be expected to be available from Q4 2018 and those who wish to see the card in action can visit AMD booth #2501 at the VMworld 2018 venue. It remains to be seen if NVIDIA is planning for a Turing-based datacenter solution sometime later this year or early 2019.

AMD Radeon Pro V340 is based on the Vega architecture. (Source: AMD)
AMD Radeon Pro V340 is based on the Vega architecture. (Source: AMD)



+ Show Press Release
static version load dynamic
Loading Comments
Comment on this article
Please share our article, every link counts!
> Notebook / Laptop Reviews and News > News > News Archive > Newsarchive 2018 08 > AMD announces the Radeon Pro V340 GPU with 32 GB HBM2 VRAM for VDI applications
Vaidyanathan Subramaniam, 2018-08-27 (Update: 2018-08-27)
Vaidyanathan Subramaniam
I am a cell and molecular biologist and computers have been an integral part of my life ever since I laid my hands on my first PC which was based on an Intel Celeron 266 MHz processor, 16 MB RAM and a modest 2 GB hard disk. Since then, I’ve seen my passion for technology evolve with the times. From traditional floppy based storage and running DOS commands for every other task, to the connected cloud and shared social experiences we take for granted today, I consider myself fortunate to have witnessed a sea change in the technology landscape. I honestly feel that the best is yet to come, when things like AI and cloud computing mature further. When I am not out finding the next big cure for cancer, I read and write about a lot of technology related stuff or go about ripping and re-assembling PCs and laptops.