Darren Crews, Sr. Principal Engineer of NPU Architecture, explains why the next-gen NPU in upcoming Intel Core Ultra processors places more emphasis on vector compute, and how transformer models have proven that multiply-accumulate often isn’t the only demand for many modern AI workloads.
Click or tap the featured link at the bottom of this video to watch the full conversation.
Click or tap the featured link at the bottom of this video to watch the full conversation.
- Category
- Intel
Be the first to comment
Up Next
Autoplay
-
01:46
Three Amazing Lunar Lake & Arrow Lake Gaming Handhelds
-
00:59
Lunar Lake: Security In the Age of AI
-
00:29
Rackspace Managed Security: The Secret Sauce of Cyber Security
-
00:59
A New Lunar Lake Gaming Handheld from MSI
-
14:33
MSI x Intel at CES 2025: Arrow Lake and Lunar Lake Gaming Hardware | Talking Tech | Intel Technology
-
55:51
Architecture All Access: Live at Lunar Lake ITT: Lunar Lake Architecture Overview
-
01:00
Phi-3 SLM on Lunar Lake Learns What You Teach It
-
00:43
Lunar Lake NUC is a Tiny but Powerful AI PC
-
00:59
Arrow Lake Laptop Built-in GPU Will Have Powerful Secret Weapon
-
00:56
Breaking down Lunar Lake’s Xe2 GPU Performance
-
12:25
How to Set Up WooCommerce Product Filters for Your Online Store
-
08:51
Best AI Content Writers in 2025: Our Top Picks
-
07:05
How to Set Up Email on Apple Mail in Minutes | Hostinger
-
14:47
How to Create a Professional Fitness Website With Ease
-
10:15
How to Host an Ark Server: Hostinger Tutorial
-
14:14
How to Build a Small Business Website | From Zero to Launch
-
17:31
10 Hobbies for Students That ACTUALLY Make You Money
-
05:49
SSH Connection Refused: 6 Methods to Fix It (Hostinger Tutorial)
-
17:01
How to Create a Business Website with AI | Fast & Easy!
-
12:56
How to Build a WordPress Website With AI