MIPS Shifts Strategy to Design AI Chips for Robotics and Autonomous Vehicles

MIPS Reinvents Itself for the AI Era

MIPS, a Silicon Valley veteran with roots dating back to the mid-1980s, has announced a significant strategic shift towards designing chips for artificial intelligence-enabled robots and autonomous vehicles 1

1

2

2

. This move marks a new chapter for the company that once competed directly with Arm Holdings in providing computing architecture.

Historical Context and Recent Developments

Founded by Stanford University Professor John Hennessey, MIPS initially focused on commercializing a nimble computing architecture 1

1

. The company's chips gained recognition for their rapid data processing capabilities in specialized applications such as networking gear and self-driving cars 2

2

. After navigating through various ownership changes and a bankruptcy, MIPS emerged in 2021 with a renewed focus on the RISC-V computing architecture, an open alternative to Arm 1

1

.

Strategic Shift to Chip Design

MIPS is now pivoting from solely licensing technology to designing its own chips, while still maintaining its licensing business 1

1

. The company's new strategy centers on three critical areas of robotics:

1. Sensing chips
2. Decision-making chips
3. Motor and actuator control chips

This approach aligns with the growing market for physical AI, which includes applications in industrial robotics and autonomous platforms 3

3

.

The Atlas Product Suite

MIPS has introduced the Atlas product suite, which includes the M8500, I8600, and S8200 solutions 3

3

. These offerings are designed to provide turnkey enablement for the three categories of computing that constitute physical AI: Sense, Think, and Act. The Atlas Explorer software platform allows customers to develop and deploy applications using MIPS Atlas subsystems in a digital twin environment, potentially accelerating time to market 3

3

.

Automotive Focus and Future Projections

Sameer Wasson, CEO of MIPS, emphasized the company's initial focus on the automotive industry. He projected that their technology would be integrated into cars by late 2027, with volume production expected in 2028 1

1

2

2

. The company aims to have select customers evaluating the M8500 real-time compute subsystem with Atlas Explorer by mid-2025, with evaluation boards available in Q4 2025 and reference silicon platforms in the first half of 2026 3

3

.

Market Opportunity and Industry Perspective

The physical AI market represents an estimated $1 trillion opportunity, according to MIPS 3

3

. Industry analysts have responded positively to the company's new direction. Jim McGregor, principal analyst at Tirias Research, noted that MIPS's multi-threaded architecture and optimized instructions are well-suited for the compute requirements of physical AI and robotics 3

3

.

Competitive Landscape

While MIPS is not directly competing with companies like Nvidia in building AI brains, it is positioning itself in the real-time control systems market with embedded AI capabilities 3

3

. This approach could potentially complement the work of other AI chip manufacturers, focusing on the specific needs of robotics and autonomous vehicles.

As MIPS embarks on this new strategy, it aims to leverage its core competencies in safety, efficient data processing, and autonomy experience to establish itself as a key player in the evolving landscape of physical AI and edge computing 3

3

.