Avicena is demonstrating individual microLED chip-to-chip links running at 14Gbit/s to reduce the power consumption of high speed links.

The LightBundle microLED technology is being shown at the European Conference for Optical Communications (ECOC) 2022 in Switzerland this week. It is based on arrays of gallium nitride (GaN) microLEDs that use the microLED display technology and can be integrated directly onto high performance CMOS ICs. Each microLED array is connected via a multi-core fibre cable to a matching array of CMOS-compatible PDs.

“We have already demonstrated LightBundle links running at less than 1pJ/bit,” said Bardia Pezeshki, founder and CEO of Avicena, “Here at ECOC 2022 we are demonstrating individual microLED links running at 14Gbit/s. Compact, low-cost interconnects using hundreds of these links can support many terabits per second.”

“We have just closed our Series A funding with a distinguished group of existing and new investors, and we will use the new funds to scale our team and build initial products for our growing family of partners and customers,” he said.

Conventional optical communications technologies developed for networking applications have been impractical for inter-processor and processor-memory interconnects due to their low bandwidth density, high power consumption, and high cost. Co-packaging existing laser sources with hot ASICs does not fit well for reliability reasons unless external laser sources (ELS) are used which increases complexity and cost.

“All of this is now changing,” says Pezeshki. “We are developing ultra-low power, high-density optical transceivers based on microLED arrays. These innovative devices leverage recent display industry advances and would have been impractical just a few years ago. Our optimized links support up to 14Gbps per lane over -40°C to +125°C temperature with excellent reliability. A LightBundle interconnect uses hundreds of parallel optical lanes connecting a microLED-based optical transmitter array to a simple CMOS-based optical receiver array over multi-core fiber cables to create low-cost multi-Tbps interconnects with up to 10 meter reach.”

The parallel nature of LightBundle is also  well-matched to parallel chiplet interfaces like UCIe, OpenHBI, and BoW, and can also be used to extend the reach of existing compute interconnects like PCIe/CXL, and HBM/DDR/GDDR memory links, as well as various inter-processor interconnects such as Nvidia’s NVLink with low power and low latency.

All material on this site Copyright © 2022 European Business Press SA. All rights reserved.