Product Description
Xilinx® Alveo™ U200 Data Center accelerator cards are designed to meet the constantly changing needs of the modern Data Center, providing up to 90X higher performance than CPUs for key workloads, including machine learning inference, video transcoding, and database search & analytics. Built on the Xilinx 16nm UltraScale™ architecture, Alveo accelerator cards are adaptable to changing acceleration requirements and algorithm standards, capable of accelerating any workload without changing hardware, and reducing overall cost of ownership.
Enabling Alveo accelerator cards is an ecosystem of Xilinx and partner applications for common Data Center workloads. For custom solutions, Xilinx’s Application Developer Tool Suite (SDAccel™) and Machine Learning Suite provide the tools for developers to bring differentiated applications to market.
Are you an app developer? If so, you might be eligible for significant discount on the U200. Register with the Accelerator Program today.
Register NowKey Features & Benefits
Fast - Highest Performance
- Up to 90X higher performance than CPUs on key workloads at 1/3 the cost
- Over 3X higher inference throughput and 3X latency advantage over GPU-based solutions
Adaptable – Accelerate Any Workload
- Machine learning inference to video processing to any workload using the same accelerator card
- As workload algorithms evolve, use reconfigurable hardware to adapt faster than fixed-function accelerator card product cycles
Accessible - Cloud <-> On-Premises Mobility
- Deploy solutions on the cloud or on-premises interchangeably, scalable to application requirements
- Applications available for common workloads, or build your own with the application developer tool
For full product specifications refer to the Datasheet
| Card Specifications | U200 | |
|---|---|---|
| A-U200-P64G-PQ-G | A-U200-A64G-PQ-G | |
| Compute | ||
| INT8 TOPs (peak) | 18.6 | 18.6 |
| Dimensions | ||
| Height | Full | Full |
| Length | ¾ | Full |
| Width | Dual Slot | Dual Slot |
| Memory | ||
| Off-chip Memory Capacity | 64 GB | 64 GB |
| Off-chip Total Bandwidth | 77 GB/s | 77 GB/s |
| Internal SRAM Capacity | 35 MB | 35 MB |
| Internal SRAM Total Bandwidth | 31 TB/s | 31 TB/s |
| Interfaces | ||
| PCI Express | Gen3x16 | Gen3x16 |
| Network Interfaces | 2x QSFP28 (100GbE) | 2x QSFP28 (100GbE) |
| Logic Resources | ||
| Look-up Tables (LUTs) | 892,000 | 892,000 |
| Power and Thermal | ||
| Maximum Total Power | 225W | 225W |
| Thermal Cooling | Passive | Active |
We’ve developed an ecosystem of Xilinx and partner solutions for most common 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. Compare how Alveo Data Center accelerator cards perform compared to traditional CPU architectures.
Data Analytics
| Vendor | Application | Acceleration vs. CPU | Accelerator Card Support | Cloud Support |
|---|---|---|---|---|
| Algo-Logic | Low Latency Key-Value Store (KVS) | 100X - 1000X | U200 | |
| Bigstream | Bigstream Hyper-Acceleration Layer for ETL, Streaming Analytics, and SQL Analytics | 2X - 30X | U200 | |
| BlackLynx | BlackLynx High Speed Search |
90X | U200 | AWS |
| InAccel | Accelerated Apache Spark MLlib | 3x-12x | U200 | AWS |
| Titan IC | Hyperion 10G RegEx File Scan | N/A | U200 | AWS |
| Vitesse Data | Deepgreen DB - Scalable MPP Data Warehouse Solution | 86X | U200 | AWS |
| Xelera | Spark-MLlib Random Forest Acceleration | 50X | U200 | OTC |
Finance
| Vendor | Application | Acceleration vs. CPU | Accelerator Card Support | Cloud Support |
|---|---|---|---|---|
| Maxeler Technologies Inc | Standard Initial Margin Model (SIMM) calculation | 10X - 50X | U200 | |
| Maxeler Technologies Inc | Real-Time Risk Dashboard | 100X | U200 | AWS |
Genomics
| Vendor | Application | Acceleration vs. CPU | Accelerator Card Support | Cloud Support |
|---|---|---|---|---|
| Falcon Computing Solutions, Inc. | Accelerated Genomics Pipelines | 10X | U200 | AWS, Huawei, Aliyun |
Machine Learning
| Vendor | Application | Acceleration vs. CPU | Accelerator Card Support | Cloud Support |
|---|---|---|---|---|
| Mipsology | Neural Network Inference for Image Classification | 100X - 500X | U200, U250 | AWS |
| SumUp Analytics | SumUp Analytics Nucleus Platform | 100X | U200 | AWS |
| Xilinx | Machine Learning Suite for Inference (Tensorflow, Caffe and MXNet) | 20X | U200, U250 | AWS |
Video and Image Processing
| Vendor | Application | Acceleration vs. CPU | Accelerator Card Support | Cloud Support |
|---|---|---|---|---|
| CTAccel | CTAccel Image Processing Accelerator | 10X | U200 | AWS |
| Skreens | Skreens Personalized Streaming Video Engine | 5X | U200 | AWS |
| Xilinx, NGCodec, VYUSync | Adaptive Bit Rate Video Transcoding Solution | 10X | U200 |
Default
Default
Title
Document Type
Date
Set Up Alveo U200 On-Premises
Download files below and follow Getting Started Guide to install.
Test Drive Alveo U200 in The Nimbix Cloud
Try NowSet Up the Alveo U200 Accelerator Card
Download the Xilinx Runtime and Deployment Shell, then follow the Getting Started Guide to install and setup your card.
1.
Download the Xilinx Runtime
The Xilinx runtime (XRT) is a low level communication layer (APIs and drivers) between the host and the card.
IMPORTANT: Please enter the following command before installing the XRT:
RedHat:
$ sudo yum-config-manager --enable rhel-7-server-optional-rpms
$ sudo yum install -y https://dl.fedoraproject.org/pub/epel/epel-release-latest-7.noarch.rpm
CentOS:
$ sudo yum install epel-release
2.
Download the Deployment Shell
The deployment shell is the communication layer physically implemented and flashed into the card.
Develop Your Own Alveo U200 Accelerated Applications
If you are a developer, download these additional files then follow the Getting Started Guide to install.
3.
Download the Development Shell
The shell interface for development is required if you are building your own applications.
4.
Download SDAccel Design Environment
The Xilinx SDAccel IDE provides a framework for developing accelerated applications.
