Uploaded October 2025 | Updated September 2026, 1 week ago
Panther Lake brings Lunar Lake efficiency and Arrow Lake performance together in a scalable architecture that enables a wide range of mobile devices. Talking Tech welcomes a cast of experts to dive deep into key elements of the architecture. Lead Client SoC Architect Arik Gihon unpacks how it scales across fabrics, memory, and a memory side cache to balance performance and efficiency; lead CPU Architect Stephen Robinson explains how we ported compute to 18A and evolved our hybrid core strategy with Cougar Cove and Darkmont; Intel Fellow Rajshree Chabukswar tackles how scheduling and power management comprehend the new cores and demands of other IP; Intel Fellow Tom Petersen explores the bigger Xe3 GPU and XeSS multi-frame generation, along with the efficient NPU5 and how we’re enabling agentic AI locally; IPU Product Marketing Manager Tomer Rider describes how integrated image processing improves quality and efficiency for video calls; and Wireless CTO and Intel Fellow Carlos Cordeiro breaks down compelling new features available with WiFi 7 and Bluetooth connectivity.
See more of Panther Lake from ITT 2025: youtube.com/playlist?list=PL8t1FdN2Tj3amQ2J4e43txyMOeD0RGu-M
Learn about the evolution of foundational tech on The Road to Panther Lake: youtube.com/playlist?list=PL8t1FdN2Tj3a8dkQruuF8ZytUaxE4EwQQ
Chapters:
00:00 – Introduction
02:03 – Scaling multiple IPS on the SoC fabric
05:11 – Flexible memory support for DDR5 and LPDDR5
08:28 – Balancing SoC performance and power efficiency
10:25 – Adding memory side cache shared across IP
11:40 – What’s new with CPU cores
12:41 – Porting the cores to 18A
14:49 – Nanocode optimizations in Darkmont E-cores
16:58 – Hybrid strategy with P-cores, E-cores & LP E-cores
19:23 – Caching across cores and system
22:06 – Combining compute from Arrow Lake and Lunar Lake
23:51 – Optimizing scheduling and power management
26:10 – Evolution of Thread Director scheduling
28:39 – Intelligently improving performance on battery
30:48 – Providing partners flexible for tuning different devices
32:53 – Challenges balancing performance and power
34:05 – Going big on graphics building up to Xe3
35:15 – Up to 12 Xe cores with enhanced XMX
36:25 – XeSS 3 adds multi-frame generation
37:54 – Neural Radiance Fields
39:27 – Improving power management for gaming
42:18 – NPU5 area efficiency & programmability
45:10 – Agentic AI you can run locally
48:58 – Quantizing models for local execution
51:26 – OpenVINO API, runtime, and Windows ML
53:20 – The importance of image and video processing
56:20 – Components of the camera system
58:24 – Dual exposure staggered HDR
59:26 – AI-based noise reduction in raw
1:02:04 – AI-powered local tone mapping
1:03:25 – Support for up to three cameras
1:04:54 – Challenges with integrated imaging & devices
1:08:46 – The new version of WiFi 7
1:09:43 – Multi-link operation
1:13:48 – Restricted target wait time
1:16:15 – Peer-to-peer channel coordination
1:19:21 – Bluetooth LE audio with Auracast
1:21:20 – Dual Bluetooth antennas
About Intel Technology:
Intel has always been at the forefront of developing exciting new technology for business and consumers including emerging technologies, data center servers, business transformation, memory and storage, security, and graphics. The Intel Technology YouTube channel is a place to learn tips and tricks, get the latest news, and watch product demos from both Intel and our many partners across multiple fields.
Connect with Intel Technology:
Visit Intel Technologies WEBSITE: intel.ly/IntelTechnologies
Panther Lake brings Lunar Lake efficiency and Arrow Lake performance together in a scalable architecture that enables a wide range of mobile devices. Talking Tech welcomes a cast of experts to dive deep into key elements of the architecture. Lead Client SoC Architect Arik Gihon unpacks how it scales across fabrics, memory, and a memory side cache to balance performance and efficiency; lead CPU Architect Stephen Robinson explains how we ported compute to 18A and evolved our hybrid core strategy with Cougar Cove and Darkmont; Intel Fellow Rajshree Chabukswar tackles how scheduling and power management comprehend the new cores and demands of other IP; Intel Fellow Tom Petersen explores the bigger Xe3 GPU and XeSS multi-frame generation, along with the efficient NPU5 and how we’re enabling agentic AI locally; IPU Product Marketing Manager Tomer Rider describes how integrated image processing improves quality and efficiency for video calls; and Wireless CTO and Intel Fellow Carlos Cordeiro breaks down compelling new features available with WiFi 7 and Bluetooth connectivity.
See more of Panther Lake from ITT 2025: youtube.com/playlist?list=PL8t1FdN2Tj3amQ2J4e43txyMOeD0RGu-M
Learn about the evolution of foundational tech on The Road to Panther Lake: youtube.com/playlist?list=PL8t1FdN2Tj3a8dkQruuF8ZytUaxE4EwQQ
Chapters:
00:00 – Introduction
02:03 – Scaling multiple IPS on the SoC fabric
05:11 – Flexible memory support for DDR5 and LPDDR5
08:28 – Balancing SoC performance and power efficiency
10:25 – Adding memory side cache shared across IP
11:40 – What’s new with CPU cores
12:41 – Porting the cores to 18A
14:49 – Nanocode optimizations in Darkmont E-cores
16:58 – Hybrid strategy with P-cores, E-cores & LP E-cores
19:23 – Caching across cores and system
22:06 – Combining compute from Arrow Lake and Lunar Lake
23:51 – Optimizing scheduling and power management
26:10 – Evolution of Thread Director scheduling
28:39 – Intelligently improving performance on battery
30:48 – Providing partners flexible for tuning different devices
32:53 – Challenges balancing performance and power
34:05 – Going big on graphics building up to Xe3
35:15 – Up to 12 Xe cores with enhanced XMX
36:25 – XeSS 3 adds multi-frame generation
37:54 – Neural Radiance Fields
39:27 – Improving power management for gaming
42:18 – NPU5 area efficiency & programmability
45:10 – Agentic AI you can run locally
48:58 – Quantizing models for local execution
51:26 – OpenVINO API, runtime, and Windows ML
53:20 – The importance of image and video processing
56:20 – Components of the camera system
58:24 – Dual exposure staggered HDR
59:26 – AI-based noise reduction in raw
1:02:04 – AI-powered local tone mapping
1:03:25 – Support for up to three cameras
1:04:54 – Challenges with integrated imaging & devices
1:08:46 – The new version of WiFi 7
1:09:43 – Multi-link operation
1:13:48 – Restricted target wait time
1:16:15 – Peer-to-peer channel coordination
1:19:21 – Bluetooth LE audio with Auracast
1:21:20 – Dual Bluetooth antennas
About Intel Technology:
Intel has always been at the forefront of developing exciting new technology for business and consumers including emerging technologies, data center servers, business transformation, memory and storage, security, and graphics. The Intel Technology YouTube channel is a place to learn tips and tricks, get the latest news, and watch product demos from both Intel and our many partners across multiple fields.
Connect with Intel Technology:
Visit Intel Technologies WEBSITE: intel.ly/IntelTechnologies
