Hardware-Adaptable Acceleration for Memory-Bound Compute

The AMD Alveo™ V80 compute accelerator brings massive hardware parallelism and memory bandwidth for workloads with large data sets. The card offers the highest logic density, network throughput, and memory bandwidth in the AMD Alveo portfolio.1

Product Advantages

Hardware-Adaptable for Memory-Bound Workloads

Features FPGA fabric to adapt the hardware to the application, coupled with HBM2e for large data sets and memory-intensive compute.

2X Logic Density and Memory Bandwidth

Delivering twice the logic density and memory bandwidth vs. the previous generation, the AMD Alveo V80 card is the most capable compute accelerator in the Alveo portfolio1.

Familiar FPGA Design Flow

Enabled by the AMD Vivado™ Design Suite for traditional FPGA flows, the Alveo V80 accelerator is coupled with a design example tailored to Alveo hardware for ease of bring-up.

Alveo V80 Infographic

Fast, Adaptable Performance for Memory-Intensive Workloads

Learn more about the breakthrough bandwidth and performance of the Alveo V80 compute accelerator, highlighting the cost savings and efficiency gains when deployed in the data center.

Meet the AMD Alveo V80 Compute Accelerator

Network-Attached, Hardware-Adaptable Acceleration

As a production board in a PCIe® form factor, the AMD Alveo V80 compute accelerator card is a faster path to production compared to designing your own PCIe card. 

Powered by an AMD Versal™ HBM device delivering 820 GB/s, the card features 4x200G networking, PCIe® Gen4 and Gen5 interfaces, DDR4 DIMM slots for memory expansion, and Mini-Cool Edge I/O (MCIO) connectors to scale across compute and storage nodes at PCIe® Gen5 speeds.

AMD Alveo V80 de-lidded image

Performance

Density & Bandwidth

1.8X Memory Bandwidth¹
810 GB/s
460 GB/s
V80
U55C
2X Logic Density (LUTs)¹
2.6 LUTs
1.3 LUTs
V80
U55C
4X Network Bandwidth¹
800 Gb/s
200 Gb/s
V80
U55C
2X PCIe® Bandwidth¹
64 GB/s
32 GB/s
V80
U55C

Accelerating Memory-Bound Applications

super computing research
High Performance Computing

Supporting custom data types with capability for 100s of nodes, the network-attached AMD Alveo V80 accelerator is suitable for a range of HPC applications, including genomic sequencing, molecular dynamics, and sensor processing.

server image
Networking

Featuring built-in cryptographic engines and adaptable hardware for custom packet processing, the AMD Alveo V80 accelerator is ideal for firewall and packet monitoring. Custom data movement also makes the card suitable for GPU clustering in data center networks.

data image
FinTech and Blockchain

For FinTech and blockchain workloads demanding massive parallelism and involving large data sets, the AMD Alveo V80 card is ideal for algorithmic trading back-testing, option pricing, and Web3 blockchain cryptography.

code over data center image with faded edge
Storage Acceleration

Ideal for compression in storage server nodes equipped with SSD storage drives, the AMD Alveo V80 accelerator allows for more effective storage capacity per server.

Orange and blue technology background circuit board and code. 3d Illustration
Hardware Emulation

For hardware emulation of ASIC designs, the AMD Alveo V80 card brings hardware flexibility and network connectivity to validate IP for communication protocols and custom IP prior to silicon tapeout.

server data image
Database Acceleration & Data Analytics

By providing low-latency processing coupled with HBM for large data sets, the AMD  Alveo V80 card allows for energy efficiency, scalability, and fast time to insights.

Partner Solution Briefs

Ciphersonic

CipherSonic Labs secures data with quantum-resistant encryption, enabling enterprises to process sensitive data without ever decrypting it using Fully Homomorphic Encryption (FHE) on the AMD Alveo V80 and Alveo U55C accelerator cards.

Dheyo

In this solution brief, Dheyo AI shows how to improve memory and security through their Agentic AI solution on the AMD Alveo V80 accelerator card.

Nera

Nera Systems enables companies to confidently innovate and collaborate on sensitive data by keeping that data private, secure, and fully under its owner’s control through Fully Homomorphic Encryption on AMD FPGAs and adaptive SoCs.

Familiar FPGA Design Flows

The AMD Alveo V80 card is fully enabled for traditional hardware developers through the Alveo Versal Example Design (AVED), available on GitHub. AVED simplifies hardware bring-up and jumpstarts development using traditional FPGA and RTL flows, based on the AMD Vivado Design Suite.

The example design is tailored for AMD Alveo hardware and provides an efficient starting point using a pre-built PCIe subsystem implemented on the Versal HBM device. It includes Alveo Management Interface (AMI) host software for control and a stress test synthetic workload (XBTEST) for easy bring-up and testing in your server of choice.

Contact Us

Sign up to receive the latest data center news or get in touch with an AMD Sales representative.

Footnotes
  1. Based on AMD comparison of the specifications of the Alveo V80 vs the Alveo U55C accelerator cards in the publicly available AMD Alveo Product Selector Guide, April 2024. ALV-13.