Intel Core Ultra 7 265KF 20-Core 3.9GHz

Intel Core Ultra 7 265KF 20-Core Desktop Processor — Socket LGA-1851 | 3.9GHz Base / 5.5GHz Boost | 125W TDP

The Intel Core Ultra 7 265KF is a Series 2 Arrow Lake desktop processor built for gamers and creators who want maximum CPU performance without paying for integrated graphics — pair it with your discrete GPU and let it fly.

20 Cores. 20 Threads. Pure Performance.

With 20 high-performance cores and 20 threads, the Core Ultra 7 265KF is engineered for demanding gaming sessions, content creation pipelines, and AI-accelerated workloads — all without compromise.

Fast Clock Speeds

A 3.9GHz base clock boosts up to 5.5GHz, delivering instant responsiveness in competitive gaming, video rendering, and heavy multitasking scenarios.

Large L2 Cache

36MB of L2 cache keeps frequently accessed data close to the cores, reducing latency and improving real-world performance across games and professional applications.

Built-In AI Acceleration

The integrated AI Boost NPU delivers 13 TOPS of dedicated AI performance, enabling Windows Studio Effects and broad AI framework support including OpenVINO, WindowsML, DirectML, ONNX RT, and WebNN.

No Integrated Graphics — Built for Discrete GPU Builds

The 265KF omits integrated graphics, making it the ideal choice for dedicated gaming and workstation builds where a discrete GPU handles all display output — often at a lower price point than its 265K counterpart.

DDR5-6400 Ready

Native DDR5-6400 (PC5-51200) support ensures your memory keeps pace with the processor, delivering high bandwidth for gaming, AI, and creative workloads.

Specifications

• Architecture: Intel Arrow Lake (Series 2)
• Cores / Threads: 20 / 20
• Base Clock: 3.9 GHz | Boost Clock: 5.5 GHz
• L2 Cache: 36MB
• Socket: LGA-1851
• Memory Support: DDR5-6400 (PC5-51200)
• Integrated Graphics: No
• NPU: AI Boost — 13 TOPS
• TDP: 125W | Max Temp: 105°C (221°F)
• Intel Virtualization Technology: Yes
• Intel Trusted Execution Technology: Yes
• AI Frameworks: OpenVINO, WindowsML, DirectML, ONNX RT, WebNN