Description
Intel Labs is pioneering the next wave of AI capabilities through neuromorphic computing and engineering. This research initiative moves beyond conventional deep-learning algorithms by co-designing specialized hardware with advanced AI software, drawing inspiration from neuroscience to address the energy demands of current AI systems. The goal is to accelerate the development of adaptive AI, making it more sustainable and efficient.
Neuromorphic computing leverages insights from neuroscience to tackle the challenges of energy-intensive AI. Intel Labs provides developers with tools to advance this field, aiming to bring neuromorphic technology to commercial applications. This research is fostering a growing community dedicated to pushing the boundaries of AI.
Hala Point represents a significant advancement, being the industry's first 1.15 billion neuron neuromorphic system designed for more sustainable AI. It features architectural improvements that deliver over 10x more neuron capacity and up to 12x higher performance compared to its predecessor. This system is a testament to Intel's commitment to scaling neuromorphic solutions.
Loihi 2 is Intel Lab's second-generation neuromorphic processor, offering up to 10x faster processing capabilities than its predecessor. It is complemented by Lava, an open-source software framework that supports various AI methods and hardware for developing neuro-inspired applications. This combination empowers researchers and developers to explore novel AI architectures.
Kapoho Point, a compact 8-chip Loihi 2 board, enables developers to scale solutions for larger problems. It can be stacked for large-scale workloads, supporting AI models with up to one billion parameters or optimization problems with up to eight million variables. This modular design facilitates the development of complex AI systems.
The Intel Neuromorphic Research Community (INRC) is a global collaborative effort uniting academic groups, government labs, and companies. This community works together to overcome challenges in neuromorphic computing and advance brain-inspired AI from research prototypes to industry-leading products. Membership is free and open to qualified groups, fostering a vibrant ecosystem for innovation.
Research utilizing Loihi 2 processors has demonstrated significant gains in efficiency, speed, and adaptability for small-scale edge workloads. Examples include Ericsson Research optimizing telecommunications AI models, a project assisting pediatric patients with neuromorphic AI, and research mimicking the sense of smell using computer chips. Intel also collaborates with institutions like Sandia National Laboratories and the National University of Singapore to explore the potential of neuromorphic computing for various applications, including robotics and large-scale computational problems.
Highlights
Co-designs optimized hardware with next-generation AI software.
Draws on neuroscience insights for AI development.
Aims to create more sustainable and energy-efficient AI.
Provides tools for developers to advance neuromorphic research.
Features Hala Point, a 1.15 billion neuron neuromorphic system.
Offers Loihi 2, a second-generation neuromorphic processor.
Includes Lava, an open-source software framework for neuro-inspired applications.
Kapoho Point board allows for scaling AI models and optimization problems.
Supports the Intel Neuromorphic Research Community (INRC) for global collaboration.
Focuses on sparse event-driven computation.
Applies brain-inspired computing principles like spiking neural networks (SNNs).
Demonstrates efficiency and speed gains for edge workloads.
Enables development of AI for sensing, robotics, and healthcare.
Use It For
- Sustainable AI
- Adaptive AI
- Robotics
- Healthcare
- Edge Computing
- Telecommunications AI
- Sensing Applications
- Optimization Problems
