I’ve used Azure, AWS, Google Cloud and other platforms for the AI workloads of D-Cube solutions. However, after some experimentation, I’ve found that a scalable 5G platform deployed on edge resources, is a great option if you want to build a high-performance and scalable solution, especially when it comes to media services in outdoor events. If you are also looking for a platform that will allow:
- Reduced service deployment complexity
- Low latency
- Security on sensitive data
- Adaptiveness to solution needs
… you should consider FLAME!
FLAME is a 5G platform
FLAME is a 5G platform which uses distributed resources connected by dynamic network routing, offering major benefits in terms of Quality of Service (QoS) and Quality of Experience (QoE) The list of features mentioned above includes only some of the reasons why the backend modules of IMRA, the Immersive Marathon Runner’s Application, were deployed at the FLAME platform, as the need was to achieve high-quality video content streaming on-demand with low latency at multiple concurrent users.
Easy to deploy
If 5G is not your expertise, that is OK. You will only need standard integration experience with commonly used technologies such as TOSCA, LXD Containers, REST etc. No need to deal with IP addresses and routing algorithms. This is the platform’s job. You are free to focus on your core
Naturally, all cloud providers such as Microsoft Azure, AWS and Google Cloud Platform can offer a vast amount of cloud computing resources for their developers. On the other hand, edge computing can alleviate some of the bandwidth requirements and provide lower latency. Especially for media services such as IMRA, the time required for the cloud to send high-quality video data to the cloud for processing and streaming it back to the device can be prohibitive. In addition to fast data transfers, AI performance at the edge is continuously improving, and enabling even the most computationally intensive AI services with minimum impact in performance.
FLAME has a lead over any cloud solution if we consider the security issues when dealing with sensitive data. Many use cases require the storage and processing of sensitive data such as people’s photos, location etc. With FLAME those data will never leave the edge, facilitating the user’s acceptance.
Adaptiveness to solution needs
Last but not least, the platform provides adaptability features that are hard to find elsewhere. The flexibility to deploy only the resources that you need and adapt them in real-time depending on the demand makes sure that you will never get overpriced for resources that you do not need. FLAME is able to scale fast, geographically and transparently to the end-user.
All in all, 5G seems the only option to eliminate latency, allowing mobile devices to connect nearly instantly. To be honest, it seems unlikely to get such low latency in a solution where the components are deployed far from the edge. This is clearly a strong point of FLAME that could tempt businesses to migrate their cloud solutions to the FLAME platform.