AMD Announces Mass-Production Of The Alveo V80 Compute Accelerator, $9495 Price Tag

AMD has announced mass production of its Alveo V80 compute accelerator, the firm's newest HPC product targeted at memory-intensive workloads.

AMD's Alveo V80 Targets Memory-Intensive HPC Workloads Through Heterogeneous Integration Courtesy of Versal HBM

The Alveo V80 was a silent announcement by AMD, but it does indeed pack a punch. This card targets catering to mid-tier workloads that are heavily dependent on memory utilization and expects to witness massive adoption in the fields of HPC, Networking, and Storage Applications. The FPGA-based accelerator looks to make strides in the industry through its competitive performance and the price tag it comes with, which we'll discuss later.

Related Story AMD Strix Halo Render Reveals Powerful Ryzen APU Design: 16 Zen 5 Cores, 40 RDNA 3+ GPU Cores, 64 MB L3 Cache

Diving into the specifications of the Alveo V80, it features AMD's 7nm-based Versal XCV80 HBM series adaptive SoC, featuring 10,848 DSP (Digital Signal Processing) slices. This SoC integrates powerful logic resources with high-bandwidth memory (HBM) to ensure fast data processing capabilities. Through the Versal HBM tech onboard, Team Red aims to provide high memory bandwidth to its consumers. With that, the Alveo V80 comes with 820 GB/s peak bandwidth, resulting from the 32 GB HBM2e DRAM (2x 16 GB stacks) equipped with the accelerator.

Regarding network and memory details, the Alveo features an onboard PCIe Gen4 x16 or PCIe Gen 5 x8 slot and a single DDR4 slot capable of equipping up to 32 GB of memory. The Alveo V80 has ample network interfaces, with 4x QSFP56 optical ports, each providing 2x 100G of interconnect bandwidth. With a power draw of up to 190W and a dual-slot form factor, AMD plans to see the Alveo V80 massively adopted in edge AI and similar environments and ensure adequate performance with the Versal architecture onboard.

AMD's Alveo V80 is priced at $9,495.00 at MSRP and is available at the firm's website. It's important to note that while the