Signal

Mount Thor, which is building cloud infrastructure for agentic AI on Apple silicon, raised $45M in a seed round and a Series A co-led by CRV and Conviction

First reported by 9to5Mac ·

The signal ●●●● Compiled by AI from 9to5Mac, Techmeme, Mount Thor Blog and TechRadar
Why you might care

The cost to rent a Mac in the cloud falls.

What happened

Mount Thor, a startup focused on building cloud infrastructure for AI agents using Apple silicon, has secured $45 million in funding. The investment includes a seed round led by Chad Byers and Offline Ventures, and a Series A co-led by CRV and Conviction. Additional participants include CSC Leasing, Enea Capital, Long Journey Ventures, Upper90, and angel investors from prominent AI companies. Mount Thor plans to leverage this funding to expand its compute footprint and build out its team, with a specific focus on creating a leading inference lab for open-source models on Apple hardware. The company offers its services through managed compute, providing bare-metal Macs or macOS VMs, and customer-managed compute, where organizations use their own Macs within Mount Thor facilities. This move targets the growing demand for macOS environments for AI agent tasks, which the company argues is now driven by agent needs rather than human users.

What it means

The substantial seed and Series A funding for Mount Thor signals a growing market demand for specialized cloud infrastructure tailored to AI agent workloads, specifically leveraging the capabilities of Apple silicon. This indicates a potential shift where the utility and performance of macOS and Apple's custom chips are becoming critical for advanced AI development and deployment, moving beyond traditional Linux-based cloud offerings. The company's focus on providing elastic compute instances of Macs at data-center scale suggests a new avenue for cloud services, catering to tasks uniquely suited to the macOS environment, such as training reinforcement learning models or running secure agent sandboxes.

This development has significant implications for AI developers and organizations that rely on or could benefit from macOS-specific functionalities. It suggests a future where access to powerful, scalable Apple hardware in the cloud is no longer a bottleneck for certain AI agent operations. As Mount Thor aims to build the best inference lab for open-source models on Apple silicon, it could accelerate the adoption and performance of these models within the Apple ecosystem, potentially influencing both AI hardware design and software development practices across the industry.

AI-written summary. May contain errors.