Intel has unveiled its next-generation client and server processor architecture based on the 18A process, ushering in the 2nm era.

Intel announced on the 10th that it will mass produce its first client SoC based on the 18A process, 'Panther Lake', and 'Xeon 6+ (codenamed Clearwater Forest)' for servers.

Panther Lake is an Intel® Core™ Ultra series 3 processor that is used in various fields such as AI PC, gaming, and edge computing. It is equipped with up to 16 P-cores and E-cores and 12 Xe GPU cores, and the CPU and graphics performance are improved by more than 50% each, and it provides AI computing performance of up to 180 TOPS. Multi-chiplet architecture provides flexibility across form factors and price points.

The Xeon 6+ for servers supports up to 288 E-cores, delivering a 17% increase in IPC and significantly improving density and power efficiency.

Optimized for hyperscale data centers and cloud environments, the product is expected to be released in the first half of 2026.

Both products are based on Intel's 2-nanometer-class 18A process, and utilize ribbon FET and power via technologies to maximize performance and energy efficiency.

Production will take place at Intel's Fab 52 in Chandler, Arizona, a key part of Intel's $100 billion U.S. manufacturing investment plan.

With this announcement, Intel sets the standard for next-generation computing platforms for the AI era, accelerating its efforts to strengthen its semiconductor supply chain and secure technological leadership in the United States.