AI coding tools write 29% of new US software code. Nvidia buys Groq for $20 billion. MCP moves to the Linux Foundation after rapid adoption.
AI Coding Tools: Production Use Hits Critical Mass
A new study published in Science reveals that AI-assisted coding has reached mainstream adoption in the United States. By early 2025, 29% of all newly written software functions relied on AI assistance, jumping from just 5% in 2022.
The research shows uneven global adoption patterns. Germany reached 23% AI-assisted code, France hit 24%, and India climbed to 20%. China and Russia lag at 12% and 15% respectively, facing barriers from government restrictions and limited access to leading language models.
Developer experience matters more than raw usage. Less experienced programmers use AI for 37% of their code, while experienced developers use it for 27%. The productivity gains of 3.6% accrue exclusively to experienced developers who leverage AI to explore new libraries and unfamiliar domains.
Stack Overflow's 2025 Developer Survey confirms the trend. 65% of developers now use AI coding tools at least weekly. The shift from autocomplete features to full agentic coding marks a fundamental change in how software gets built.
AI Processing: Nvidia's $20 Billion Groq Acquisition
Nvidia announced its largest acquisition ever, purchasing AI chip startup Groq's assets for $20 billion. The deal brings Groq's language processing unit technology and CEO Jonathan Ross into Nvidia's operations.
Groq developed processors that run large language models 10 times faster than traditional GPUs while using one-tenth the energy. Ross previously helped create Google's tensor processing unit, making him a key hire for Nvidia's expanding AI infrastructure play.
The acquisition follows Nvidia's launch of the Rubin platform at CES 2025. The platform combines six new chips into an AI supercomputer system. The Rubin GPU delivers 50 petaflops of compute for AI inference, with a third-generation Transformer Engine that includes hardware-accelerated adaptive compression.
Nvidia claims the Rubin platform cuts inference token costs by 10 times and reduces GPUs needed to train mixture-of-experts models by 75% compared to the Blackwell platform. Microsoft plans to deploy hundreds of thousands of Rubin systems in its Fairwater AI data centers.
High-bandwidth memory emerged as a critical bottleneck. Micron Technology estimates the HBM market will grow from $35 billion in 2025 to $100 billion in 2028. Leading chip designers pack massive amounts of HBM into their AI accelerators to prevent memory bandwidth from limiting GPU performance.
Standards and Protocols: MCP Joins Linux Foundation
Anthropic donated the Model Context Protocol to the Agentic AI Foundation in December 2025, marking a major milestone for AI interoperability. The protocol reached 97 million monthly SDK downloads and 10,000 active servers in its first year.
MCP became the standard way for AI agents to connect with external data sources and tools. OpenAI, Google, Microsoft, and other major platforms now support MCP natively. The protocol works alongside Google's Agent2Agent protocol, which handles communication between different AI agents.
Google launched Agent2Agent in April 2025 with over 50 industry partners. The protocol enables AI agents from different vendors to collaborate on tasks, share state information, and coordinate workflows. Both protocols moved under Linux Foundation governance, preventing vendor lock-in.
The Linux Foundation formed the Agentic AI Foundation with founding members including Anthropic, OpenAI, Block, Microsoft, Google, AWS, and Bloomberg. The foundation aims to create vendor-neutral standards for autonomous AI systems that plan and execute tasks independently.
Security challenges emerged alongside rapid adoption. Research in July 2025 found nearly 2,000 MCP servers exposed to the internet without authentication. The June 2025 spec update addressed authorization concerns by implementing OAuth Resource Server classification and requiring RFC 8707 Resource Indicators.
Experience the Future with Dremio
The AI landscape changes fast. Data teams need tools that keep pace.
Dremio's semantic layer and Apache Iceberg foundation let you build AI-ready data products. The platform handles optimization automatically. You focus on insights, not infrastructure.
Ready to see agentic analytics in action? Start your free trial today and experience the autonomous lakehouse.
Top comments (0)