Hardware Adaptable

Optimize the platform for the workload
and adapt to evolving algorithms and application requirements.

Real-Time Performance

Massive parallelism for high throughput and performance/watt, as well as low latency for real-time applications.

Accessible to Any Developer,
On-Premise or Cloud

Supports traditional FPGA flows and software development flows; accessible for on-premise or cloud deployment.

Data Center Applications

Data Analytics FinTech High Performance Computing Media Acceleration Infrastructure Acceleration

AMD Alveo Accelerator Card Portfolio

General-Purpose Compute Accelerators

Among our most versatile accelerator cards in the portfolio, our compute accelerators targeting workloads including analytics, HPC, packet monitoring, network switching, storage, and algorithmic trading.

Alveo™ V80

Highest logic density, HBM bandwidth, DSP compute, and network throughput in the Alveo portfolio.

Alveo™ U55C

Flagship HBM accelerator for compute intensive workloads in a FHHL form factor.

Alveo™ U50

Slim HHHL HBM accelerator for memory-bound compute.

FinTech Accelerators

Our purpose-built FinTech cards address the unique requirements of financial markets by converging low latency networking with adaptive compute—ideal for a breadth of workloads including trading, risk analysis, and market data distribution.

Alveo™ X3522PV

Low latency performance in a slim HHHL form factor.

Alveo™ UL3422

Slim form factor ultra-low latency trading with purpose-built FPGA technology.

Alveo™ UL3524

Ultra-low latency trading with purpose-built FPGA technology.

Network Accelerators

Offload server CPUs from infrastructure tasks in the data center with our network accelerators, delivering line-rate performance for a breadth of infrastructure workloads, such as virtual switching, IPsec, and computational storage.

Alveo™ U45N

2x100G with FPGA flexibility, supporting OpenNIC reference design for ease of development.

Media Accelerators

Video transcoding and media processing for high channel density at low latency. Alveo media accelerators are enabling a new era of interactive media services that demand a high volume of concurrent streams.

Alveo™ U30

Low latency live streaming supporting H.264, H.265 codecs.

Alveo™ MA35D

Ultra-low latency streaming supporting AV1, H.264, and H.265.

AI Inference Accelerators 

The AMD VCK5000 development card features AI engines for breakthrough compute– based on Versal adaptive SoCs.

AMD VCK5000 Versal™ Development Card

Up to 145 INT8 TOPs, 2x100G connectivity, broad application with customizable FPGA logic.

Developers

FPGA Development Flows

Traditional FPGA design flows are supported on AMD Alveo accelerator cards using the AMD Vivado™ Design Suite. Developers can easily get started with a pre-validated base design that maps directly to Alveo hardware and access pre-optimized building blocks in the Vivado IP catalog.​

Accessible in the Cloud

FPGA acceleration is also available in the cloud to quickly develop proof-of-concept. Enterprises can leverage cloud options for evaluation, production deployment, or employ a hybrid strategy.

AWS

Based on up to eight AMD Virtex UltraScale+ VU9P devices, the F1 instance supports Vivado and Vitis flows for HW and SW developers.

Microsoft Azure

Scalable to up four AMD Alveo™ U250 accelerators, the NP instance enables a breadth of workloads using the Vitis Unified Software Platform.

VMaccel

Supporting the majority of AMD Alveo cards, VMaccel is an ideal evaluation vehicle for the Alveo portfolio using Vitis flows.

Server Support

Our network of OEM qualified servers a broad selection of deployment options for our Alveo accelerator cards. View our catalog to find a provider and configuration that meets your needs.

Contact Us

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