The AMD Alveo™ U55C high performance compute card provides optimized acceleration for workloads in high performance computing (HPC), big data analytics and search, financial computing, computational storage and machine learning.
Featuring powerful Virtex™ XCU55 UltraScale+ FPGA from AMD, the Alveo U55C card packs in high bandwidth memory (HBM2) and 200Gbps of high-speed networking into a single slot, small form factor card, and is designed for deployment in any server.
The U55C accelerator is paired with the new standards-based API-driven clustering solution from AMD and equipped with C/C++ programmability. This solution empowers a broad spectrum of customers with large-scale compute workloads to realize the performance advantages of FPGAs at 1000+ node scale, easily implementing HPC clustering using their existing data center infrastructure and network, with no proprietary hardware required.
The U55C card supports PCI Express® (PCIe®) Gen3x16 or dual Gen4x8, contains 16 GB of High-Bandwidth Memory (HBM2) at 460GB/s of bandwidth, and Ethernet networking capability with dual QSFP28 connectors capable of 2 x 100 Gb/s.
Enabling Alveo accelerator cards is an ecosystem of AMD and partner applications for common data center workloads. For custom solutions, the Application Developer Tool Suite (Vitis™ environment) and Machine Learning Suite from AMD provide the frameworks for developers to bring differentiated applications to market.
The Alveo Programmable Cable is required for development access on Alveo U55C cards. The cable provides micro-USB support from a host PC to the Alveo U55C maintenance connector.Alveo Accessories
Built for Performance & Efficiency
Adaptable – Accelerate Any Workload
Accessible - Cloud <-> On-Premises Mobility
|Board Specifications||Alveo U55C High Performance Compute Card
|Look-up Tables (LUTs)||1,304K|
|HBM Memory Capacity||16 GB|
|HBM Total Bandwidth||460 GB/s|
|Internal SRAM Capacity||43 MB|
|Internal SRAM Total Bandwidth||35 TB/s|
|PCI Express||Gen3x 16, 2 x Gen4x 8|
|Network Interfaces||2x QSFP28|
|Vitis Developer Environment||Yes|
|Vitis Unified Software Platform||Gen3x16 XDMA|
|Vivado Design Suite||Yes|
|Power and Thermal|
|Maximum Total Power||150W|
|Target Workloads||big data analytics and search, financial computing, computational storage, and machine learning|
We’ve developed an ecosystem of AMD and partner solutions for compute and data intensive workloads. Alveo Data Center accelerator cards can deliver dramatic acceleration across a broad set of applications and are reconfigurable to provide an ideal fit for the changing workloads of the modern data center.
The preferred optimal design flow for targeting the Alveo Data Center accelerator card uses the Vitis™ software platform. Steps to deploy and develop using Vitis are given below. Long-time FPGA designers might want to use traditional design flows, such as RTL or HLx. This flow does not require installing the Vitis platform.
For development using RTL and HLx, follow these steps: