Arm vs Cerebras Systems

Arm CEO Confident $2B AI CPU Revenue Goal Achievable

Arm Holdings CEO Rene Haas told CNBC's Jim Cramer on September 16, 2026, that the company is increasingly confident it can meet Wall Street's $2 billion revenue target for its new in-house data center CPU, the AGI CPU. The demand for the chip has been strong, but investors have focused on Arm's ability to secure sufficient manufacturing capacity to convert that demand into revenue. Haas noted that confidence in achieving the $2 billion figure has grown from May to July and further strengthened by September. The AGI CPU marks a significant expansion of Arm's business model, moving from licensing chip designs to selling complete chips. Arm first disclosed $2 billion in demand in May, double its initial $1 billion outlook, but maintained a conservative official outlook until supply confidence improved. Shares have declined about 45% from their June high after a parabolic run earlier in the year.

• Arm Holdings CEO Rene Haas expressed increased confidence in meeting a $2 billion revenue goal for its AGI CPU data center chip.
• The $2 billion target is double the $1 billion demand Arm initially outlined in March 2026.
• Arm's stock fell 10% after maintaining a $1 billion revenue outlook due to supply concerns in May.

Arm introduces new AI-native compute platform built for agentic AI and mobile graphics

Arm unveils Arm AI Portal, expands AI infrastructure with AGI CPU and Neoverse CSS N4, and brings ecosystem together for physical AI.

Arm introduces new AI-native compute platform built for agentic AI and mobile graphics

Arm announced a new AI-native compute platform for agentic AI and mobile graphics, expanded AI infrastructure with AGI CPU and Neoverse CSS N4, and unveiled the Arm AI Portal to accelerate optimized AI apps.

Cerebras and Compute Nordic Finland Announce New 165 MW AI Data Centre in Mikkeli, Finland

Cerebras Systems and Compute Nordic Finland announced a new 165 MW AI data centre in Mikkeli, Finland. The company also unveiled the CS-4, claiming up to 30 times faster than GPU-based solutions, and announced a partnership with Callosum for ultra-low-latency heterogeneous agentic inference.