Loihi Chip

A Chip that can Smell!!

“Need is the mother of invention!”

The growing requirements of robots to undertake and assist humans in all their undertakings has encouraged the origination of a chip that replicates mammals in the power of senses too. Wait for a minute and think, what if your PC smells the odour and also gives you an idea about the source of it!! A PC with a nose?

This chip “Loihi” isn't your typical silicon chip but rather is neuromorphic, developed to work the same way as a brain with digital circuits that communicate like neurons. It’s more of a digital recreation of a nose.It uses an asynchronous spiking neural network (SNN) to implement adaptive self-modifying event-driven fine-grained parallel computations used to implement learning and inference with high efficiency.

The chip was formally presented at the 2018 Neuro Inspired Computational Elements (NICE) workshop in Oregon.The chip is named after the Loihi volcano as a play-on-words - Loihi is an emerging Hawaiian submarine volcano that is set to surface one day.

It’s Origin

As a result of over a decade of studying olfactory bulb circuitry in rodents and trying to figure out essentially how it works, the researchers from Intel and Cornell University have developed a neuromorphic chip that can identify and learn the smell. While Thomas Cleland, a professor of psychology at Cornell University, was studying the biological olfactory system in mammals by measuring the electrical activity seen in their brains on smelling different odours, he configured a set of algorithm describing the use of Very-large-scale-integration(VLSI) systems containing electronic analog circuits to mimic neuro-biological architecture present in the nervous system. It is like a human brain with digital circuits that communicate like neurons.

Announced in September 2017, Loihi is predominantly a research chip meaning performance characteristic are not guaranteed. This is Intel's 5th chip in the neuromorphic category. The first three chips were early internal test chips while the fourth is a 10 nm wafer-level probe study. The chip was initially tested and simulated using FPGAs. Actual silicon implementations arrived in late November.

How does this Algorithm work?

This is based on rapid learning and robust recall in neuromorphic olfactory circuit. Identification of odorant samples under noise, based on the architecture of the mammalian olfactory bulb and implemented on the Intel Loihi neuromorphic system. As with biological olfaction, the spike timing-based algorithm utilizes distributed, event-driven computations and rapid (one shot) online learning. Spike timing-dependent plasticity rules operate iteratively over sequential gamma-frequency packets to construct odour representations from the activity of chemosensory arrays mounted in a wind tunnel. Learned odorants then are reliably identified despite strong destructive interference. Noise resistance is further enhanced by neuromodulation and contextual priming. Lifelong learning capabilities are enabled by adult neurogenesis. The algorithm is applicable to any signal identification problem in which high-dimensional signals are embedded in unknown backgrounds.

Implementation on the Loihi neuromorphic system

Combining the methods like Mitral cell implementation, Granule cell implementation, Excitatory synaptic plasticity and Inhibitory synaptic plasticity customizes integrated circuits that model biological neural computations, typically with orders of magnitude greater speed and energy efficiency than general-purpose computers. These systems enable the deployment of neural algorithms in edge devices, such as chemosensory signal analyzers, in which real-time operation, low power consumption, environmental robustness, and compact size are critical operational metrics. Loihi is fabricated in Intel’s 14-nm FinFET process and realizes a total of 2.07 billion transistors over a manycore mesh. Each Loihi chip contains a total of 128 neuromorphic cores, along with three embedded Lakemont x86 processors and external communication interfaces that enable the neuromorphic mesh to be extended across many interlinked Loihi chips.

Demonstrations and testing of chip

The chip, powered by the neural algorithm, is able to identify and even learn the neural pattern of odours based on inputs from an array of sensors. The chip was able to identify odours even when their pattern was 80% different from the one it had learned originally.In a demonstration, the chip was able to learn and recognize the scents of 10 hazardous chemicals based on data they fed to it, by 72 chemical sensors. The chip, based on the electrical responses, was able to mimic the biological olfaction system and learn each smell. It was then able to identify each smell, even in the presence of interfering odours.

Loihi based neuromorphic system

• Kapoho Bay (2 chip, 262k neurons)

Kapoho Bay is a USB stick form factor that incorporates 1 or 2 Loihi chips. Announced on Dec 6, 2018, Kapoho Bay includes a USB host interface and a DVS interface for neuromorphic sensors such as a camera. With 2 chip Kapoho Bay has 256 neuromorphic cores with 262,144 neurons and 260,000,000 synapses.

• Nahuku (32 chip, 4M neurons)

Nahuku is a scalable Arria10 FPGA expansion board. Intel uses the Nahuku board as the framework for building larger systems. The Nahuku board comes in multiple configurations from 8 to 32 chips. Those chips are organized as 16 chips in a 4 by 4 grid mesh on both sides. With 32 chips there is a total of 4,096 neuromorphic cores incorporating a total of 4,194,304 neurons and 4,160,000,000 synapses. With the Nahuku board, an FPGA host is connected to a set of conventional sensors such as actuators as well as neuromorphic sensors such as a DVS camera or a silicon cochlea. The board communicates with a standard "super host" CPU which can be used to send commands to the board and to the management core on the chips themselves.

Future Implementations of this neuromorphic chips

This may prove to be a milestone in the field of robotics. The Intel’s Loihi neuromorphic chip can smell hazardous chemicals and also exhibit the property of self learning of new smell. This “electronic nose systems" technology has many potentials in the field of chemistry to identify compounds on the basis of their peculiar smell and can also be very useful for assistants at chemical labs and industry. You never know one day using this AI, people would be able to generate all the senses which would could sense even better than the human’s original ones.