Before the Pandemic, experts had conflicting views about the impact that edge computing might have on cloud computing giants. Some even believed that edge computing could replace centralized clouds in a manner much like the way that personal computers replaced mainframes. The reality has proven to be quite different. Edge computing, which brings applications closer to data sources, has become more of a complement to cloud services than a replacement. This trend was predicted by Deloitte’s David Linthicum, and has become especially true as AI-based applications explode.
The AI boom is leading to more intelligent devices. The complex training models these devices require are often better suited for cloud-based processing due to their computational demands. This trend argues in favor of edge computing solutions, and it presents a significant opportunity for infrastructure-as-a-service providers like Amazon Web Services, Microsoft Azure, and Google Cloud. The edge computing market is expected to grow over 30% annually in the coming years, fueled by the growth of artificial intelligence (AI) applications.
Cloud providers are well-positioned to dominate the AI-at-the-edge market. They can provide essential services like configuration management, data management, security, and operations, which are more efficiently managed from a central cloud. Despite concerns about latency, most edge use cases don't require real-time responsiveness, and cloud providers have invested in infrastructure and partnerships to meet various edge computing needs.
In light of the growing importance of data in artificial intelligence, the leading research and advisory company, Gartner, has identified several key trends that it expects to shape the future of data science and machine learning (DSML). These trends include:
1. Cloud Data Ecosystems: Data ecosystems are transitioning from standalone software to cohesive cloud-native solutions. Gartner predicts that by 2024, 50% of new cloud system deployments will be based on integrated cloud data ecosystems rather than individual solutions. Organizations are advised to evaluate these ecosystems based on their ability to handle distributed data challenges and integrate data from various sources.
2. Edge AI: The demand for Edge AI is rising as it enables real-time data processing at the source, providing immediate insights and meeting data privacy needs. Gartner forecasts that over 55% of deep neural network data analysis will occur at edge systems by 2025, compared to less than 10% in 2021. Organizations are encouraged to identify applications for edge environments near IoT endpoints.
3. Responsible AI: Responsible AI focuses on ethical and business considerations when adopting AI, aiming to make AI a positive force for society. Gartner predicts that by 2025, 1% of AI vendors will concentrate pretrained AI models, making responsible AI a societal concern. Organizations are advised to adopt a risk-proportional approach to AI, ensuring vendors manage risks and comply with obligations.
4. Data-Centric AI: Data-centric AI emphasizes a shift towards building AI systems based on data rather than just models and code. Techniques like AI-specific data management and synthetic data creation are addressing data challenges. Gartner predicts that by 2024, 60% of data for AI will be synthetic, simulating real-world scenarios and reducing the reliance on actual data.
5. Accelerated AI Investment: Investment in AI is rapidly increasing across many industries, with over $10 billion projected to be invested in AI startups relying on foundation models by the end of 2026. A Gartner survey found that recent excitement around technologies like ChatGPT led 45% of executives to boost their AI investments. Many organizations are exploring generative AI, with 19% already in pilot or production mode.
While there are still challenges to adoption of edge solutions, such as latency-sensitive applications, cloud providers are actively focusing on the edge. Companies like AWS, Google, and Microsoft are expanding their presence at the edge and offering tools for model training. AWS has built a network of more than 450 globally dispersed points of presence for low-latency applications. Google LLC has 187 edge locations and counting. Microsoft has 192 points of presence. All three cloud providers are also striking deals with local telcos to bring their clouds closer to the edge.
The market for edge computing is not expected to be dominated by a single provider. Cloud providers will face competition from telecom carriers, content delivery networks, and specialized silicon manufacturers. The edge computing landscape is evolving with various players contributing to its growth and development.