Amazon Web Services and Google Cloud have built a new way for their clouds to talk to each other directly. The service combines AWS Interconnect—multicloud with Google Cloud Cross-Cloud Interconnect to give customers private, dedicated links instead of going over the public internet.
This setup targets companies that already run apps and data across multiple clouds and want cleaner, simpler connectivity. It is now in preview on AWS in five regions, with Google Cloud live as the first partner and Azure planned for 2026.
How the new multicloud network works
The old way needed customers to order physical cross-connects, deploy hardware, and coordinate with multiple network teams. That often meant weeks or months before traffic could reliably flow between AWS and Google Cloud.
The new offer takes that pain away by turning multicloud links into a managed, cloud-native experience. Users can log into their preferred console or use APIs, request dedicated bandwidth, and bring up connections in minutes.
Key building blocks include:
- AWS Interconnect—multicloud as the AWS-side managed gateway for other cloud service providers.
- Google Cloud Cross-Cloud Interconnect is part of Google’s broader Cross-Cloud Network architecture.
- Open API specifications on GitHub that other providers can implement to plug into the same model.
Focus on speed, security, and reliability
Both companies say reliability and security are core design goals. The architecture uses four-way (quad) redundancy across separate interconnect facilities and routers so that a single failure does not break service.
Traffic between AWS and Google Cloud edge routers is protected with MACsec encryption, which secures data at Layer 2 over the dedicated links. Continuous monitoring on both sides aims to spot and fix issues before they impact customers.
For customers already using services like AWS Transit Gateway, AWS Cloud WAN, or Amazon VPC, the new product is meant to slot in with minimal extra work. On the Google side it extends the Cross-Cloud Network vision of giving one consistent way to connect on-prem, Google Cloud, and other clouds.
Why this matters for enterprises and AI
Multicloud use has grown as organizations mix best-of-breed tools, meet regulatory demands, or spread risk across providers. But complex, DIY networking made this strategy harder, especially for global companies with many regions.
This joint service could matter most in a few areas:
- Data platforms and analytics: moving large datasets between AWS and Google Cloud with lower latency and predictable bandwidth.
- AI and machine learning: grounding AI models in data that may live on different clouds, such as Salesforce data and other SaaS platforms integrating through these links.
- Hybrid and regulated workloads: building private, high-speed paths that support compliance and performance needs while staying multicloud.
Salesforce already highlighted how AWS Interconnect—multicloud—helps it build private bridges to Google Cloud for Salesforce Data 360, letting customers tie AI and analytics to trusted data regardless of where it lives.
Open standard and what comes next
A notable part of the move is the open specification that governs how these interconnects work. AWS has published API specs on GitHub so other cloud providers and partners can adopt the same model, rather than relying on one-off, proprietary integrations.
AWS plans to extend Interconnect—multicloud to Microsoft Azure from 2026 onward, expanding the same pattern beyond Google Cloud. If more providers join, multicloud networking could shift from a custom project to a standardized capability that customers simply turn on when needed.
(Source: cloud.google, aws.amazon)
#AWS #GoogleCloud #Multicloud #CloudNetworking #AI
