Microsoft fully commits to Gen AI: A comprehensive rundown of all the announcements made at Ignite.

Mon Nov 20, 2023 - 10:15am GMT+0000

Microsoft’s annual Ignite conference, hosted in Seattle this year, was brimming with exciting announcements and updates concerning the company’s vision and advancements in generative AI technology.

These announcements encompassed a wide range of innovations, including the introduction of new AI chip hardware and software, security enhancements, and open-source initiatives. These efforts underscore Microsoft’s dedication to establishing leadership in the rapidly evolving realm of emerging technologies.

Here are some key highlights from the event:

Microsoft introduced “Mirroring,” a feature integrated into Microsoft Fabric, which facilitates the replication of external databases, even in proprietary formats. This innovation empowers users to seamlessly consolidate their data within Microsoft Fabric, resulting in cost savings. The replication technology is already operational for Azure Cosmos DB, Azure SQL DB, Snowflake, and Mongo DB customers, with additional data sources slated for inclusion by 2024. Unifying data remains crucial for advancing sophisticated Gen AI capabilities.

Microsoft bolstered its infrastructure capabilities with the introduction of in-house Maia and Cobalt chips designed for enterprise AI. These chips offer efficient, scalable, and sustainable computing power, enabling organizations to leverage the latest breakthroughs in cloud and AI technologies. This development represents the final piece of the puzzle in Microsoft’s mission to provide flexible infrastructure systems, combining its proprietary hardware and software with partner-delivered solutions, tailored to diverse workload requirements. Maia, an AI-centric GPU, is optimized for cloud-based training and inference for generative AI workloads, while Cobalt is an Arm-based chip designed for general-purpose workloads.

Nvidia unveiled its AI Foundry service for Microsoft Azure, featuring new Nemotron 3.8B models. This move strengthens the collaboration between Nvidia and Microsoft, allowing enterprises and startups to develop customized AI applications on the Azure cloud, including those capable of harnessing enterprise data through retrieval augmented generation (RAG).

Microsoft elevated AI data security to new heights with Purview upgrades. These enhancements empower Copilot users on Microsoft 365 to exert control over the data accessible to the AI coding assistant, automatically categorize sensitive data in responses, and implement compliance controls related to LLM usage.

Microsoft wholeheartedly embraced Gen AI with the introduction of Security Copilot and Defender XDR. These new cybersecurity solutions are designed to identify, detect, and respond to threats that often elude current detection and response systems. Microsoft’s approach leverages generative AI to uncover threats and disseminate this information across its applications, copilots, extended detection and response (XDR) systems, the cloud, and hybrid clouds in real-time. Gen AI forms the cornerstone of Microsoft’s comprehensive security strategy, reflecting the urgency of their enterprise customers to combat these undetected attacks.

Microsoft unveiled expanded Copilot capabilities at Ignite, specifically aimed at enhancing sales and service operations. These AI-driven features, integrated deeply across the Microsoft portfolio, seek to boost sales productivity and enhance customer service experiences within organizations. Users will find it easier to integrate Microsoft 365 with sources of sales and service data, surpassing previous integration capabilities.

[Update: The following news, while not part of the Ignite event, was exclusively reported by VentureBeat on the same day] Microsoft and Google have joined forces on “OneTable,” an open-source partnership aimed at supporting the OneTable project. This collaboration has the potential to reshape the landscape of cloud data lakes for years to come, emphasizing the importance of consistently formatted and reliable data for all AI projects.