Join top executives in San Francisco July 11-12 to hear how they are integrating and optimizing their AI investments for success. learn more
Enterprises are increasingly gaining a competitive advantage by deploying artificial intelligence (AI) with distributed hybrid cloud architectures.
This is due to two factors. First, more data is being generated at the edge than ever before. In fact, Gartner predicts that 50% of enterprise-generated data will be processed outside the traditional data center or cloud by 2025, and a recent global survey found that 78% of IT decision makers % see moving their IT infrastructure to the digital edge as a priority. Future-proof your business.
Second, moving large data sets to AI training infrastructure engines in central locations for processing means companies spend valuable time and money. Moreover, compliance and privacy regulations often mandate that processing and analysis of AI data remain in the country of origin, further justifying the spread of workloads across multiple countries. .
Let’s delve into three different industry use cases where distributed AI is helping organizations reduce costs, meet regulatory needs, and achieve new technological advances.
Join us July 11-12 in San Francisco. A top executive shares how she integrated and optimized her AI investments and avoided common pitfalls for success.
Gain real-time retail insights while reducing costs
Many large retailers are finding competitive advantage by leveraging distributed digital infrastructure strategies. They use what IDC recently identified as an increasingly popular AI deployment strategy. This means developing AI at the core, such as in the cloud or regional data centers, deploying AI inference models at the edge, and then retraining the models with new regional data to fit your application.
For example, a retailer using a distributed hybrid cloud model would initially send in-store camera feeds and inventory control data to a colocation metro data center to build regional AI models and leverage federated AI methods. may integrate regional models. These optimized AI models are then deployed to store locations to gain insight into inventory, employee shift management, shopper buying propensity predictions, ad placement recommendations, low/predictive Run AI model inference for latency.
Deploying the AI inference engine from one metro data center location is more cost effective than maintaining and maintaining these servers at every retail location. This decentralized AI infrastructure will enable the retailer to process and analyze insights quickly at her one regional location, ultimately increasing revenue.
Maintaining Privacy and Compliance in Video Surveillance
According to UNCTAD, most countries around the world (71%) have enacted laws governing privacy and data protection. Distributed data management and AI architectures play a key role in ensuring organizations achieve compliance.
For example, a large property management company with sites in several metropolitan areas around the world could deploy distributed AI to hundreds of security cameras around the world by deploying AI wherever data was collected. Leverage architecture to maintain compliance with local privacy regulations. Complying with local privacy laws by having centralized facilities in the various countries in which the company operates, by sending data to another country that may not have the same compliance regulations as the data originating do not violate.
This model not only achieves privacy and data usage compliance, but also reduces costs by hosting the AI inference stack in a single metro location instead of each facility. We also process motion detection data on-site at each of our hundreds of locations.
Realization of automated driving with the latest local information
Self-driving cars enabled by advanced driver assistance systems (ADAS) cannot address certain use cases without an AI infrastructure. ADAS requires AI to make decisions about how vehicles interact with their surroundings, especially when interacting with vulnerable road users such as cyclists and pedestrians.
The amount of data that test vehicles generate to train AI models is enormous. For Level 2 and 3 ADAS (where vehicles can adjust speed, brake, and make decisions based on their environment), that’s 20 TB to 60 TB per vehicle per day. AI enables connected vehicles to collect and process these massive data sets from test vehicles faster and more cost-effectively than using traditional infrastructure.
A distributed AI infrastructure is defining the next generation of vehicle mobility and autonomy. For example, connected cars leverage HD maps to provide information about signs and roads to the car. But what if there is a construction zone or road hazard at night? , can communicate the hazard to all vehicles in the area.
go with the flow of data
Nothing feels the pull of data more than AI. To get the most out of their AI infrastructure, organizations should assess the value of deploying them centrally, regionally, or locally. Doing so saves time, money and valuable latency speeds.
Doron Handel is Head of Global Business Development at Equinix.
data decision maker
Welcome to the VentureBeat Community!
DataDecisionMakers is a place for data professionals, including technologists, to share data-related insights and innovations.
Join DataDecisionMakers for cutting-edge ideas, updates, best practices, and the future of data and data technology.
You might consider contributing your own article!
Read more about DataDecisionMakers