The Cloudcast: The Cloudcast #334 - The Future of Edge Computing
"... talks with Derek Collison (@derekcollison, Founder and CEO at Synadia Communications, board member of CNCF, former CTO of VMware, Architect of Cloud Foundry, technical director at Google, TIBCO,etc.) about the future of edge computing, the impact of AI/ML on edge systems, and how Telcos and open source communities will evolve with edge computing opportunities."
Doing this computing closer to the edge of the network lets organizations analyze important data in near real-time – a need of organizations across many industries,"
Edge computing - Wikipedia
Interesting "edge" use-cases are "cloud functions" and Machine Learning inference.
ML models can be created by training on cloud, and then deployed to edge for faster response time.
That is exactly what AWS already offers with AWS Greengrass - Amazon Web Services.
Optionally combined with customized HW, i.e. optimized for ML and/or IoT, it really becomes a new platform. AWS Greengrass – Run AWS Lambda Functions on Connected Devices | AWS News Blog
Introducing AWS Greengrass @ SlideShare
Greengrass is software (bring your own hardware); Manage from cloud console or API;
price $1.5/year/device.
Microsoft also provides a similar integrated solution:
Azure IoT Edge | Microsoft Azure | Pricing - IoT Edge | Microsoft Azure
Azure IoT Hub is required for the secure management of devices and services deployed to the edge via Azure IoT Edge. Pricing—IoT Hub | Microsoft Azure (based on number of messages/month)"There is no charge for using Azure IoT Edge. Azure IoT Edge allows you to run multiple Azure services on the edge. These Azure services running on IoT Edge will be billed according to their specific pricing."
Containers are a very good technology for deploying and updating software,
and even small computers like Raspberry Pi it can run both Docker and Kubernetes.
"Edge" container management is not standardized (yet), but that could change over time...
SCaLE 13x Derek Collison NATS A new nervous system for distributed cloud platforms - YouTube
GopherCon 2014 High Performance Systems in Go by Derek Collison - YouTube
derekcollison/nats-go: NATS client for Go @ GitHub
Apcera | Interview with its Founder & CEO - Derek Collison
Our Apcera Enterprise Solutions Defined | Apcera
Interesting "edge" use-cases are "cloud functions" and Machine Learning inference.
ML models can be created by training on cloud, and then deployed to edge for faster response time.
That is exactly what AWS already offers with AWS Greengrass - Amazon Web Services.
Optionally combined with customized HW, i.e. optimized for ML and/or IoT, it really becomes a new platform. AWS Greengrass – Run AWS Lambda Functions on Connected Devices | AWS News Blog
Introducing AWS Greengrass @ SlideShare
Greengrass is software (bring your own hardware); Manage from cloud console or API;
price $1.5/year/device.
Microsoft also provides a similar integrated solution:
Azure IoT Edge | Microsoft Azure | Pricing - IoT Edge | Microsoft Azure
Azure IoT Hub is required for the secure management of devices and services deployed to the edge via Azure IoT Edge. Pricing—IoT Hub | Microsoft Azure (based on number of messages/month)"There is no charge for using Azure IoT Edge. Azure IoT Edge allows you to run multiple Azure services on the edge. These Azure services running on IoT Edge will be billed according to their specific pricing."
Containers are a very good technology for deploying and updating software,
and even small computers like Raspberry Pi it can run both Docker and Kubernetes.
"Edge" container management is not standardized (yet), but that could change over time...
SCaLE 13x Derek Collison NATS A new nervous system for distributed cloud platforms - YouTube
GopherCon 2014 High Performance Systems in Go by Derek Collison - YouTube
derekcollison/nats-go: NATS client for Go @ GitHub
Apcera | Interview with its Founder & CEO - Derek Collison
Our Apcera Enterprise Solutions Defined | Apcera
No comments:
Post a Comment