TECH – Intel has unveiled Hala Point, the most powerful neuromorphic supercomputer to date, mimicking the structure and efficiency of the human brain. Leveraging over 1,150 Loihi 2 chips, the system simulates 1.15 billion artificial neurons and up to 128 billion synaptic connections—an achievement described as the world’s largest brain-like computing system.
Designed to deliver high efficiency at low power, Hala Point operates at a mere 2,600 watts, albeit boasting up to 20 quadrillion operations per second, or 20 peta‑ops. According to Intel’s Neuromorphic Computing Lab, this new architecture is 50× faster and more than 10× larger than its predecessor, Pohoiki Springs, while consuming far less energy than traditional CPU/GPU-based AI systems.
Installed at Sandia National Laboratories in New Mexico, the system is roughly the size of a microwave oven chassis yet supports 140,544 neuromorphic processing cores, along with 2,300 embedded x86 processors for auxiliary tasks . Capable of scaling workloads over massive parallel networks, it promises breakthrough performance in machine learning and large-scale simulation.
Read More: Germany Turns Cockroaches into Spy Robot Swarms
Sandia engineers emphasize the importance of the system’s tight integration of memory, processing, and communication—an architecture inspired by biological neural networks, where computation and data flow are decentralized and tightly coupled. Hala Point’s neuron capacity is equivalent to that of an owl brain or a capuchin monkey cortex—but scaled to deliver 20× faster processing than those of the human brain (as estimated computationally).
Applications for Hala Point span from continuous real-time learning, logistics and scientific modeling to advanced AI operations including real-time analytics, deep neural net training, and smart city infrastructure planning. With energy-per-operation efficiency exceeding 15 trillion TOPS/W, the system offers sustainable scaling paths for AI workloads.
As Intel and Sandia push this neuromorphic technology further, Hala Point signals a turning point in AI infrastructure. By emulating brain-like computation at scale and energy efficiency, it presents a compelling alternative to today’s resource-intensive GPU clusters—and poses new architectural possibilities for future AI systems.
The arrival of Hala Point marks a milestone in neuromorphic computing and sheds light on a path toward smarter, more sustainable AI—and stronger convergence between biological models and silicon engineering.
