Intel has announced the development of a neuromorphic system with 1,150 million neurons, using its Loihi 2 processors to advance AI technology. Known as Hala Point, this research prototype is deployed at Sandia National Laboratories and aims to support brain-inspired AI analysis. The system can perform 20 quadrillion operations per second with greater efficiency.
Hala Point consists of 1.152 Loihi 2 processors on an Intel 4 process node, supporting 1.15 billion neurons and 128 billion synapses. It includes 140,544 neuromorphic processing cores and 2,300 x86 processors for auxiliary calculations. The system offers high memory, inter-core, and inter-chip communication bandwidth.
Intel believes Hala Point’s capabilities can enable advancements in real-time continuous learning for AI applications like scientific problem-solving, logistics, smart city management, language models, and AI agents. Compared to its predecessor, the system shows improved performance and efficiency for real-time processing tasks like video and voice.
Sandia National Laboratories researchers will utilize Hala Point to solve scientific computing problems in various fields. Intel plans to share the system with research collaborators as part of a new family of large-scale neuromorphic research systems.