GPU Technology Conference Munich 9-11 October 2018
ADLINKs’ Simon Collins to give presentation – ‘Some of the challenges in architecting large-scale distributed systems’
ADLINK is a global provider of leading Edge Computing solutions, offering a unique combination of industrial hardware, advanced data connectivity, industry leading partner software and vendor-agnostic cloud compatibility for end-to-end industrial IoT (IIoT) solutions at the Edge.
At this years’ GPU technology conference in Munich ADLINK Technology will also demonstrate the latest AI training platform ALPS-4800 plus their M200-JT2 Inference Platform and many more.
In addition, Simon Collins, Senior Product Manager, ADLINK Technology will present a close look at some of the challenges in architecting large-scale distributed systems, such as those found in Internet of Things applications. ADLINKs’ presentation will propose, through use of an established open-standard middleware, how to solve those challenges with a highly scalable data-centric architecture, allowing AI practitioners to concentrate on their applications, and decouple different parts of the system, allowing for far greater flexibility and evolutionary upgrades. Simon Collins’ session will propose a simple but robust method of data distribution for large-scale IoT deployments. The examples will highlight how to deploy an inferencing model to many edge nodes, such as the NVIDIA Jetson TX2, and how to implement a data model that best represents the Thing data being processed at the edge for consolidation, reporting and potentially collaboration and control.
Attendees will learn how a peer-peer publish/subscribe messaging technology based on data topics to facilitate collection of initial in-situ data, distribution of inferencing models, load-sharing between “worker nodes”, collation of inferencing results from many nodes to a central “command”, and collation of corner-case data to facilitate iterative updates to the trained model.
New ADLINK products in the GPU accompanying show:
• Real-time Video Analysis for Smart City
ADLINK is showcasing in the GPU accompanying show their new M200-JT2 Inference Platform. The adoption of AI is picking up speed and bringing intellig Simon Collins, Senior Product Manager, ADLINK Technology ence to different industries. To deliver intelligence at the edge, ADLINK’s M200-JT2 Inference Platform integrates NVIDIA® Jetson™ TX2 and ADLINK partner’s AI inference algorithms. The M200-JT2 can detect vehicles, identify vehicle types, and recognize moving direction of vehicles. The M200-JT2 can analyze traffic flow at an intersection in real time for traffic management optimization. The M200-JT2 can also be used in tracking vehicles, improving smart bus services, and strengthening security surveillance for smart cities.
• ADLINK’s ALPS- 4800 AI Training Platform, a carrier-grade, highly scalable platform
The new Artificial Intelligence (AI) nowadays is not only an academic subject, but is moving fast towards the real world with applications in facial recognition, robotics, revolutionary analytics, disease prevention and smart city constructions. All the groundbreaking scientific progress calls for acceleration in machine-learning (ML) and deep-learning (DL) training, and the increasing adoption of GPUs will satisfy the thirst for tremendous computing power. By leveraging ADLINK’s leadership in edge computing, the ALPS-4800 AI Training Platform is a carrier-grade, multi-purpose, and highly scalable edge platform enabling operators and their customers to quickly build and enhance services with NVIDIA’s advanced GPU technologies.
Integrating a server node, a PCIe expansion box with PCIe switching, ADLINK’s ALPS-4800 AI Training Platform has a high capacity to support a combination of up to eight NVIDIA GPUs depending on the needs of the application. As a configurable edge platform that supports different workloads and demands, the Platform can support multiple topologies and bandwidths between GPUs and CPUs with simple cable routing adjustments. In addition, Infiniband support allows it to be easily scaled up to multiple GPU clusters. Thanks to its hot-swappable and front-accessible design, it also promotes efficient serviceability while delivering optimal performance.
• ADLINK’s NEON-J Industrial Camera
The first AI industrial camera to integrate the NVIDIA® Jetson™ TX2, image sensor, and I/O control in a compact chassis serving machine vision applications at the edge.
Addressing problems beyond current machine vision algorithms, the NEON-J utilizes cutting-edge NVIDIA GPU technology and ARM Cortex-A57 processors, enabling robust Deep Learning inference, empowering defect inspection and object classification requirements in manufacturing environments.
A neural model optimized by NVIDIA tools like Digits or TensorRT can be deployed on the NEON-J directly, easing development of user-specific machine vision solutions. The NEON-J provides an Ethernet port for GigE camera connection, 4x digital input, 4x digital output, 1x communication port, and VGA output for maximum integration.
For more info on above products or to set up press appointments
please contact Knud Hartung, email email@example.com