It is now widely acknowledged that artificial intelligence (AI) is fueling a wave of mergers and acquisitions (M&A) in the technology sector. As analyzed earlier this month, large-scale AI-related deals, combined with a merger-friendly regulatory environment under the current U.S. administration, have pushed the total value of announced tech M&A transactions this year to over $543 billion—the highest since 2021.
According to Morgan Stanley bankers, the driving force behind this year’s M&A activity, as well as deals in the coming years, revolves around a central theme: companies are building or restructuring their foundational technology infrastructure for the AI era, spanning critical areas such as software, databases, cloud services, and semiconductors.
The two primary segments of M&A activity are data-centric firms and semiconductor manufacturers. Over the past three months, a flurry of data-related deals has been completed. Morgan Stanley advised on three such transactions: facilitating a $11 billion proposed acquisition of data-streaming leader Confluent by IBM, orchestrating the merger of data startups Fivetran and dbt Labs, and assisting data management firm Veeam Software in its $1.73 billion purchase of data security company Securiti.
Businesses are focusing on data utilization and management to gain an edge in the AI race. Rohan Mehra, Managing Director and Co-Head of Global AI Investment Banking at Morgan Stanley, noted, "All players within and around the data ecosystem are working to define the core requirements of AI-era data architecture and outline their technological roadmaps."
Take IBM’s proposed acquisition of Confluent as an example. Founded in 2014 and publicly listed in 2021, Confluent is a leader in real-time data analytics software. IBM’s strategic rationale lies in its anticipation that future AI systems will heavily rely on continuously flowing real-time data.
Clearly, many companies share IBM’s perspective. Sources reveal that at least one other publicly traded firm and a private equity player have expressed interest in acquiring Confluent. Bankers point out that the M&A market often exhibits a "domino effect"—once a landmark deal is struck, it tends to trigger similar transactions across the industry.
Mark Edelstein, Chairman of Global Semiconductor Investment Banking at Morgan Stanley, recently advised on the sale of Celestial AI to chipmaker Marvell Technology, a deal valued at up to $5.5 billion. He emphasized that AI is also driving major semiconductor M&A.
Edelstein, who has been involved in the semiconductor industry since the 1980s—first as a research analyst and later as an investment banker—has witnessed multiple consolidation waves, including tracking companies like Nvidia. He highlights that the sector is now at a critical inflection point, with AI development shifting focus from training large language models to deploying them in real-world applications.
He stresses that this transition imposes two key demands on chips: higher energy efficiency and reduced "memory-compute latency," which minimizes data transfer delays between processors and memory chips.
Marvell’s acquisition of Celestial AI aims to address speed bottlenecks. Celestial AI’s core technology uses photonics for data transmission, offering faster speeds compared to traditional copper-based electrical signaling. This could significantly enhance Marvell’s competitiveness in the data center chip market. Meanwhile, companies like d-Matrix and Mythic are accelerating R&D on new chips designed to shorten data transfer distances between memory and computing units.
The AI boom ignited by ChatGPT is reshaping the tech landscape, though this transformation will take time. Edelstein observes, "This revolution has already created winners and losers across tech and the broader economy—and we’re still in its early stages."