Cutting through the noise on AI PCs with Intel Core Ultra

What changes for battery life, security, and real-world productivity when AI moves onto the laptop

Sponsored Post AI PCs are everywhere right now, along with plenty of big claims. In this Register Hot Seat, Intel's Tom Pieser joins us to explain what the company's Core Ultra 200V processors are designed to deliver for business users, and where on-device AI makes practical sense today. Watch the full Hot Seat on demand.

The conversation starts with the fundamentals IT teams care about: power efficiency, performance, and security. Tom sets out Intel's headline goal: better performance at around half the power consumption compared with the prior generation, with a stated uplift in CPU performance and a significant jump in integrated graphics. He also outlines the platform's AI compute capability, and why Intel believes that matters as more workloads run across CPU, GPU and NPU.

Battery life is a major theme. Rather than treating it as a vague promise, the Hot Seat covers what "all-day" can mean in practice, including how usage patterns and IT builds can affect results. Tom points to benchmark figures pre-deployment, then contrasts that with testing under more realistic enterprise conditions, such as long Teams calls with additional effects enabled.

There is also a useful discussion of locally-run generative AI, versus generative AI in the cloud. Running workloads on the device instead of sending data to the cloud. Running generative workloads such as image generation locally, without uploading content externally offers benefits including lower latency, stronger privacy, reduced dependence on connectivity, and potential cost control depending on how services are licensed.

Security and manageability get proper attention too. The video covers Intel's work with partners on local AI-based threat detection, and why offloading some scanning to GPU or NPU can improve responsiveness for users. It also touches on how AI PCs could support more proactive

Sponsored by Intel.