A naukri.com initiative
Insider
3d
161
Image Credit: Insider
AMD's CTO says AI inference will move out of data centers and increasingly to phones and laptops
- Mark Papermaster, CTO of AMD, foresees AI inference transitioning from data centers to edge devices like smartphones and laptops, offering a market share opportunity against Nvidia.
- The shift towards AI inference, particularly in the realm of edge computing, is seen as a lucrative development by AMD.
- Papermaster anticipates the emergence of an AI 'killer app' within the next three to six years.
- AI workloads heavily shifting from training to inference is advantageous for AMD, as inference plays a pivotal role in generating outputs like answering queries or producing images.
- AMD executives recognize the potential to seize market share from Nvidia amidst the focus on inference computing.
- With an increasing demand for AI computing across various devices, AMD aims to align its direction towards meeting the evolving needs in this domain.
- Use cases for AI computing in edge devices such as laptops and phones revolve around local, immediate, and low-latency content creation.
- Applications like real-time translation and content creation are expected to become more seamless and efficient due to advancements in AI capabilities on edge devices.
- The future of handheld AI may see a majority of inference tasks being performed at the edge, with transformative applications evolving rapidly in the next few years.
- Ongoing innovations focus not only on enhancing accuracy and capabilities but also on optimizing efficiency in AI models, ensuring continuous advancement in AI computing technology.
Read Full Article
9 Likes
For uninterrupted reading, download the app