Advanced Computing in the Age of AI | Friday, March 29, 2024

IBM Opens Up with Multi-Cloud Manager, AI Tools 

shutterstock 611605280

With upwards of 85 percent of companies using more than one cloud provider, the shift to multi-cloud deployments driven by emerging open-source platforms has prompted infrastructure vendors to begin offering management tools for migrating and integrating private and public cloud applications with on-premise systems.

Among them is IBM, which this week rolled out a cloud management suite based on open-source tools to harmonize multi-cloud deployments from on-premise to public clouds. Tuned to the IBM private cloud that is based on the Kubernetes container orchestrator, the manager extends that capability to link different cloud services. It includes a dashboard interface to manage Kubernetes-based distributed applications.

Separately, the company (NYSE: IBM) also released a platform designed to boost AI adoption by addressing application development using different tools while tackling concerns about how AI applications make decisions. The company said its open approach can even help detect and correct bias in AI applications.

IBM also released new cloud security tools as part of its open-source push.

The company said Monday (Oct. 15) its multi-cloud manager can integrate workloads from Amazon Web Services (NASDAQ: AMZN), Microsoft Azure (NASDAQ: MSFT), Red Hat (NYSE: RHT)

and others running on the Kubernetes orchestrator. The dashboard would give developers visibility into Kubernetes-based applications across different clusters and clouds.

Industry watchers note that the primary challenge to multi-cloud adoption has been integration of on-premises with public cloud platforms. Hence, IBM and other vendors have increasingly been turning to agnostic tools like de facto-standard Kubernetes to move beyond early, fragmented multi-cloud deployments.

IBM said it multi-cloud manager will be available by the end of October.

The company also moved deeper into the enterprise AI sector this week with the release of what is billed as the first open platform for building and running AI applications. Along with its own Watson AI tools, the AI OpenScale platform also handles applications built with Microsoft’s AzureML, the TensorFlow machine learning framework, AWS SageMaker or other tools.

The growing emphasis on open source platforms reflects the requirement for greater interoperability among many moving parts if users are going to achieve scale. IBM likens to the current status of AI to that of the Internet 25 years ago: A promising technology lacking in trust.

“A lot of work had to be done behind the scenes before we had faith in the [Internet],” the company noted.

The open source AI effort also seeks to address the bias that is inherent in machine learning algorithms. The company argues that many AI models go unused “because businesses don’t yet trust them. They don’t know if they contain bias. They don’t know how they reach their conclusions. They can’t trace their logic.”

Hence, the IBM AI platform promises nothing less than the ability to understand and therefore trust AI models to help make business decisions while at the same time augmenting the skills needed to manage AI applications.

Completing the open-source push is an IBM cloud platform for cyber-security applications that uses AI to analyze security data shared across different clouds and among “previously unconnected tools,” the company said.

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).

EnterpriseAI