• Intel develops the world’s largest neuromorphic computer system, named Hala Point.
  • Hala Point contains 1,152 Loihi 2 processors, supporting up to 1.15 billion neurons and 128 billion synapses.
  • The system achieves up to 50 times faster performance compared to traditional GPUs while using 100 times less energy.

Intel has unveiled Hala Point, the world’s largest neuromorphic computer system. This innovative hardware stack promises groundbreaking advancements in AI capabilities and computational efficiency.

Unprecedented scale and design

Hala Point integrates 1,152 Loihi 2 processors, boasting a remarkable capacity to support up to 1.15 billion neurons and 128 billion synapses. See also: LARUS launches LARUS ONE partnership framework.

The system’s architecture, comprising 140,544 neuromorphic processing cores, reflects Intel’s commitment to pushing the boundaries of AI research and development. See also: AI workload volatility raises data centre power waste.

Also read: Intel reveals details of new AI chip to take on Nvidia

Exceptional performance and efficiency

Leveraging its technology, Hala Point achieves up to 20 quadrillion operations per second, surpassing the computational capabilities of traditional GPU-based systems.

Despite its performance, Hala Point consumes only a fraction of the energy required by conventional hardware, marking a significant leap in computational efficiency. See also: OpenAI IPO turns compute procurement into a market test.

Also read: Intel expects to ship 40 million AI PCs in 2024

Potential for advancement and impact

Intel envisions Hala Point as a catalyst for future AI innovation. Intel said Hala Point “could enable future real-time continuous learning for AI applications” like AI agents, large language models and smart city infrastructure management.

“We hope that research with Hala Point will advance the efficiency and adaptability of large-scale AI technology.” said Mike Davies, director of the Neuromorphic Computing Lab at Intel Labs. See also: Pure DC brings German biomethane to Dublin data centre.

Besides, Google DeepMind explores NeuroAI, emphasizing AI learning from memory, not just data accumulation. IBM’s NorthPole semiconductors mimic human brain processing on one chip.