Lightelligence to Demonstrate Hummingbird Optical Network-on-Chip
Photonic computing company Lightelligence will demonstrate Hummingbird, an Optical Network-on-Chip (oNOC) processor designed for domain-specific artificial intelligence (AI) workloads, at Hot Chips 2023, to be held at Stanford University from 27-29 August. Maurice Steinman, Lightelligence’s Vice President of Engineering, will present “Hummingbird Low-Latency Computing Engine” on 29 August at 11:30 a.m. P.D.T.
Introduced in June, Hummingbird utilizes advanced vertically stacked packaging technologies to integrate a photonic chip and an electronic chip into one single package serving as the communications network for data centres and other high-performance applications. It is the first in a family of products that utilize Lightelligence’s oNOC platform that improves computing performances by enabling innovative interconnect topologies via silicon photonics.
Hummingbird’s waveguides propagate signals at the speed of light and utilize an all-to-all data broadcast network to each core on a 64-core domain-specific AI processor chip for latency and power reduction advantages over traditional digital interconnect solutions. With 64 transmitters and 512 receivers, Hummingbird provides a framework to implement a variety of dense optical network topologies. Unlike digital networks, Hummingbird’s oNOC technology increases density scaling by enabling interconnect topologies that would be otherwise unrealizable.
In oNOC, power and latency are virtually unaffected by distance, making the technology ideal for developing new and more robust topologies that do not rely on nearest neighbour communication. Hummingbird’s oNOC topologies enable higher utilization of compute power even in a single electronic IC configuration due to more efficient communication. With oNOC, mapping workloads to hardware becomes easier and provides greater freedom to select the right topology for the computing task.
Hummingbird’s electronic and photonic ICs are co-packaged and integrated into a PCIe form factor ready for installation in industry-standard servers. Coupled with the Lightelligence Software Development Kit (SDK), machine learning and AI workloads can be optimized to take full advantage of the oNOC. oNOC and Hummingbird IP can also be customized for other unique workloads and applications.
