The Big Denominator EDGE is a general-purpose application development and deployment framework. Its simple and intuitive interface facilitates the rapid development, testing, deployment, and administration of high performance applications. It is targeted to those needing customized solutions but who may lack the skills and/or investment required for full application development.
- Automatically exploit modern multi-core processor architectures and SIMD (single-instruction, multiple data) operations;
- Handle both synchronous and asynchronous design patterns;
- Seamlessly scale from a single computer to a full network cluster; and
- Provide both low latency (individual transaction time) and high throughput (volume).
EDGE applications are structured as directed graphs – individual components (models) which, when stitched together, interact to form a complete system. Analogously, a car is not merely a single entity but also a composite of many subsystems (e.g., electrical, hydraulic, fuel). Each of these subsystems can be viewed as a graph, further decomposable into individual parts. Thus, applications are simply groups of small, independent, reusable processing modules (a.k.a. “graph nodes”). The Big Denominator runtime layer – the DGE (Directed Graph Engine) – stitches the components together, handles input and output data, and manages memory and communication.
The EDGE API and design process empowers customers to be effective high performance computing (HPC) developers – channeling their unique insight and subject matter expertise into highly efficient and effective HPC solutions – without getting bogged down in computer science jargon. The entire API for each node to interact with other nodes consists of three methods: REQUEST, SEND, and VIEW. All production and consumption of data – regardless of type, technology, source or destination (intra-/inter-server – is addressed through this gateway.
Lastly, EDGE applications are at home in streaming, real-time environments as well as batch processing and a mix of both. EDGE applications can collaborate with Spark and other streaming, “Big Data” utilities as well as directly interface with many databases and messaging systems. Moreover, EDGE API allows you to create your own custom external interfaces.
Contact us for a test drive.