Azure Updates: AI Infrastructure; VMs; Cloud for Healthcare

Microsoft marked the middle of October by shedding more light on its Azure AI infrastructure. Senior Director Yvonne Muench discussed ways to unlock data with AI offerings in the commercial marketplace. Highly reliable data is an essential part of AI models, although models also depend on security, governance, data integration, storage, processing, and analysis. Muench gave examples of third party offerings to support AI initiatives such as Apache Kafka and Apache Flink on Confluent Cloud, Datadog’s monitoring system, ElasticSearch, or the Matillion Data Productivity Cloud. 

Matthew Lundgren, Microsoft’s Chief Scientific Officer highlighted the capabilities of Cloud for Healthcare to create comprehensive multimodal models.  A variety of AI models are no available for Cloud for Healthcare including MedImageInsight to classify images, MedImageParse for image segmentation, and CXRReportGen for analyzing chest x-rays.

According to General Manager Mike Hulme, there is a substantial opportunity for companies to modernize existing apps with AI. Microsoft has developed a variety of modernization resources. It recommends customers look to examples like the Application Modernization Checklist developed by Forrester. 

CVP Arun Kishan announced the general availability of Cobalt-based VMs. The VMs run on a 64-bit Arm-based Cobalt 100 CPU, designed in-house by Microsoft. The VMs include the Dpsv6-series and Dplsv6-series as well as memory-optimized Epsv6-series VM series. Currently, the VMs are available in a wide variety of regions like Canada Central, Central US, and Switzerland North. “The general availability of Azure Cobalt 100-based VMs marks the beginning of a new era in Azure’s infrastructure. With our custom silicon program, we are delivering exceptional price performance and power efficiency to our customers. We are excited to see the impact of these innovations on our customers’ businesses and we look forward to bringing even better solutions to our customers in the future,” Kishan wrote.