- AWS and Google Cloud teamed up to make it drastically easier to provision cross cloud connectivity using an open API
- The importance of multi-cloud connectivity ran through several conversations at re:Invent, including with Oracle and Clarifai
- AI seems to be the catalyst prompting cloud companies to solve the multi-cloud problem that's persisted for years
AWS RE:INVENT, LAS VEGAS – AI was unavoidable at re:Invent this year, dominating not just the keynote stage but also sessions and the expo floor. But there was one non-AI announcement that was decidedly a Big Deal and that was the launch of AWS Interconnect.
In a nutshell, AWS teamed with Google Cloud to develop an open API that allows them to offer a managed way for customers to provision dedicated cross-cloud connectivity in minutes using a cloud console. As IDC analyst Taranvir Singh put it in a LinkedIn post "From operations perspective, this simplifies enterprise data exchanges for data intensive AI workloads and distributed applications without [a] bespoke connectivity patchwork."
Though AWS is initially launching with Google Cloud, it noted plans to bring Microsoft Azure into the mix “later in 2026.”
It’s not yet clear if Google Cloud plans to link up to Azure as well. Rob Enns, Google’s VP of Engineering for Cloud Networking, told Fierce that “in theory” Microsoft could connect to Google Cloud in addition to AWS using the open API but “we haven’t talked specifics.” But for what it's worth, Google has been all aboard the multi-cloud train for a few years now with its Cross Cloud Network, so we wouldn't be surprised if it eventually happens.
Zooming out, Sid Nag, President and Chief Research Officer at Tekonyx, told Fierce the collaboration between AWS and Google Cloud “signals where the industry is heading: toward cloud-native, multi-provider networks where private, high-speed connectivity is available instantly, securely and without specialized networking expertise. It’s a massive win for customers and a blueprint for the future of multi-cloud.”
The new multi-cloud moment
Indeed, while this was the highest-profile announcement, the multi-cloud thread ran through several conversations Fierce had at re:Invent. Multi-cloud, it seems, is once again having a bit of a moment.
Oracle was at the show talking up Oracle Database@AWS, which is basically Oracle hardware running inside AWS data centers. Announced last year and opened for general availability in July, the product is expanding to 20 additional AWS regions over the next 12 months, with availability in three of those – Tokyo, Frankfurt and Ohio – slated for the very near term.
Oracle also has its databases deployed in 31 Microsoft Azure regions and 12 Google Cloud regions.
Multi-cloud capabilities also popped up in a conversation with Clarifai, a company that provides a platform for companies to deploy computer vision and other AI models across different cloud environments.
But why is multi-cloud back in focus? You can chalk it up to AI.
When speaking with Enns about the Interconnect announcement, the first thing we asked is why it took so long to introduce these kinds of capabilities. The question seemed fair given cloud customers have been struggling with cloud silos for years.
“The industry had to get to the point that cloud providers were getting asked by their customers to work together,” he said. “I feel that’s where we got to.”
The answer to what prompted customers to start asking came from Oracle’s VP of Multi-Cloud Nathan Thomas. Asked why on earth customers would want Oracle hardware in AWS data centers, Thomas explained that companies have historically put “their most important data in their Oracle databases.” As they now work to develop and deploy AI, they’re finding “all my most important data isn’t participating in my AI pipeline that I’m developing inside of AWS.”
AI, then, isn’t just a product but also a catalyst for a change enterprises have been clamoring for.
So, where do things go from here? According to Nag, the open API that Google Cloud and AWS published marks a “major step toward an open, interoperable cloud ecosystem. By enabling other providers and partners to adopt this standard, the industry moves closer to true multi-cloud portability without sacrificing performance or security.”