Amazon Web Services and Google Cloud have jointly launched a new multicloud networking architecture designed to simplify high-speed connectivity between cloud environments. The initiative aims to reduce deployment time from days to minutes, providing enterprises with a resilient, secure, and interoperable foundation for modern applications.
Amazon Web Services (AWS) and Google Cloud have introduced a jointly engineered multicloud networking service that allows enterprises to quickly connect workloads across cloud platforms. The launch represents a significant shift in how cloud providers approach interoperability, offering customers a managed, cloud-native path to high-speed cross-cloud connectivity.
AWS described the new approach as easing a long-standing challenge in multicloud deployments: navigating complex, multi-layer network configurations across different providers. According to AWS VP of Network Services Robert Kennedy, the collaboration reflects a “fundamental shift in multicloud connectivity,” with both companies aligning on a unified architecture.
A Simpler Multicloud Path
The service integrates AWS Interconnect – multicloud with Google Cloud Cross-Cloud Interconnect, enabling private, high-bandwidth links between the two platforms. Unlike traditional do-it-yourself multicloud designs, where organisations manually stitch together networks, routing layers, and security policies, this managed model is configured in minutes.
Customers can now privately link AWS environments such as Amazon VPC, AWS Transit Gateway, and AWS Cloud WAN to Google Cloud without building custom routing or managing intermediate network hops. During preview, bandwidth starts at 1 Gbps, scaling up to 100 Gbps at general availability.
Performance, Redundancy, and Security
The jointly developed system prioritises reliability and resilience. Quad-redundant connections provide protection against facility, edge-router, and software failures. The underlying physical links between AWS and Google Cloud are secured with MACsec encryption, offering line-rate, always-on data protection.
Organisations adopting the service gain:
- High-speed data movement across clouds
- Automated redundancy across multiple failure domains
- Private connections without traversing the public internet
- Faster deployment, reducing setup time from days to minutes
Salesforce is among the early adopters testing the architecture.
AI Workloads Accelerating Multicloud Demand
The rollout comes as enterprises shift AI workloads across multiple cloud providers to access specialised accelerators, region-specific models, and diverse AI agents. A Forbes survey cited by Google indicates the majority of organisations expect AI adoption to increase reliance on multicloud networking, particularly as training clusters and inference systems span different vendors.
Rising demand has led major cloud providers, including Amazon, Microsoft, and Alphabet, to invest billions in infrastructure upgrades. The timing also aligns with ongoing concerns about cloud reliability. The announcement followed an AWS outage on October 20 that disrupted services at apps such as Snapchat and Reddit, with estimated economic losses of up to $650 million.
Open Specification and Expansion Plans
The service is being published under an open specification, enabling other cloud and network providers to implement compatible architectures. AWS plans to extend support to Microsoft Azure in 2026.
Preview availability begins across five AWS regions, including:
- N. Virginia
- Oregon
- London
- Frankfurt
AWS remains the world’s largest cloud provider, generating $33 billion in Q3 revenue, more than double Google Cloud’s $15.16 billion.
What This Means for Enterprises
The new multicloud approach reduces operational overhead for organisations managing distributed applications across multiple clouds. It also supports architectures built for AI, analytics, financial systems, and regional compliance requirements. With a shared standard emerging between two major cloud providers, enterprises gain a more predictable, resilient networking environment for future workloads.
