Former GitHub CEO launches $60M startup, SMIC warns of idle AI data centers, and Google ships an MCP server for its developer docs.
AI Coding Tools: Ex-GitHub CEO Bets Big on Agent Code Management
Former GitHub CEO Thomas Dohmke launched Entire on February 10 with $60 million in seed funding at a $300 million valuation. Felicis led the round, calling it the largest seed investment ever for a developer tools startup. Y Combinator CEO Garry Tan, Datadog CEO Olivier Pomel, and Microsoft's M12 fund also backed the company.
Entire tackles a growing pain point: developers now manage fleets of AI coding agents that produce code faster than any human can review. Its first open-source product, Checkpoints, logs the prompts and context behind every AI-generated code change. The tool supports Claude Code and Gemini CLI at launch. "Our manual system of software production was never designed for the era of AI," Dohmke said in his press announcement.
Apple also updated Xcode to version 26.3 this week, adding support for Anthropic's Claude Agent and OpenAI's Codex inside the IDE. Developers can now use advanced reasoning models directly inside Apple's development environment. The update lets teams generate code from natural language inputs and catch errors in real time.
These launches reflect a broader shift. AI coding tools are moving beyond autocomplete into full agent workflows. The question has shifted from "can AI write code?" to "who manages all the code AI writes?" Entire, Snowflake, and Apple all answered that question differently this week.
AI Processing: SMIC Warns AI Data Center Rush Is Outpacing Demand
China's largest chipmaker sounded an alarm on February 11. SMIC Co-CEO Zhao Haijun told analysts that companies are racing to build 10 years of data center capacity in just two years. "As for what exactly these data centers will do, that hasn't been fully thought through," Zhao said. He also warned that the memory chip market is in "crisis mode", with companies overbooking orders amid a global supply crunch.
The numbers back up his concern. Moody's Ratings projects AI infrastructure spending will top $3 trillion over the next five years. In 2026 alone, Alphabet, Amazon, Meta, and Microsoft plan to spend roughly $650 billion combined on capital expenditures.
SMIC reported 2025 revenue of $9.3 billion, up 16.2% year over year. But wafer shipments from smartphone and consumer electronics makers are getting squeezed by the surge in AI chip orders. Net profit rose 39% to $685.1 million, yet both full-year and Q4 profits missed analyst estimates. SMIC's Hong Kong shares dropped 2.2% on the news.
The HBM (high-bandwidth memory) crunch adds pressure. Zhao said tight HBM supply will persist for years because new capacity takes time to build and qualify. Data centers consume an estimated 70% of all memory chips produced in 2026, leaving other sectors short.
A Deloitte outlook published this week estimates global semiconductor sales will reach $975 billion in 2026. AI chips now drive roughly half of that total revenue but represent less than 0.2% of all chip units sold. The report flags a stark risk: the industry has put nearly all its growth behind AI, and non-AI markets like automotive and consumer electronics remain flat.
Standards and Protocols: Google Launches Developer Knowledge MCP Server
Google released the Developer Knowledge API and MCP server on February 4 in public preview. The tool gives AI coding assistants direct access to Google's official developer documentation for Firebase, Android, Google Cloud, and more. Google re-indexes all docs within 24 hours of a service update.
The MCP server works with popular assistants and IDEs. Developers connect it through a simple config file and an API key from Google Cloud. Two core functions power the API: SearchDocumentChunks finds relevant doc snippets, and BatchGetDocuments retrieves full page content. Google plans to add structured content like code samples before general availability.
This launch builds on a busy period for the MCP ecosystem. Google also proposed adding gRPC transport to MCP, a move that helps enterprises already running gRPC across their services. The proposal addresses a real friction point: forcing teams to rewrite services or run translation proxies.
On the security front, the Coalition for Secure AI (CoSAI) released its MCP Security white paper in late January, cataloging nearly 40 threats across 12 categories. The paper treats MCP servers like any critical infrastructure, not just another API. It calls for zero-trust principles, sandboxing, and end-to-end traceability for every agent request.
Cisco donated Project CodeGuard to CoSAI on February 9. The open-source framework embeds security rules directly into AI coding workflows, guiding agents to produce safer code from the start. Meta also joined CoSAI as a Premier Sponsor on February 3, adding weight to the coalition's growing roster of backers including Google, IBM, Microsoft, and NVIDIA.
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)