▾ G11 Media Network: | ChannelCity | ImpresaCity | SecurityOpenLab | Italian Channel Awards | Italian Project Awards | Italian Security Awards | ...
InnovationOpenLab

Avicena Advances microLED and Photo Detector Arrays to Enable the World’s Lowest-Power AI Scale-up Optical Interconnects

#innovation--Avicena, the leader in next-generation microLED interconnects, today announced a major milestone at SuperCompute 2025 (SC25): its LightBundleTM microLED-based links now operate at 4Gbps p...

Immagine

Latest Avicena LightBundle technology delivers industry-leading Tx power of 80fJ/bit addressing next generation AI infrastructure requirements

ST. LOUIS: #innovation--Avicena, the leader in next-generation microLED interconnects, today announced a major milestone at SuperCompute 2025 (SC25): its LightBundleTM microLED-based links now operate at 4Gbps per lane with transmitter currents as low as 100µA per LED. At this current, each link closes at a raw BER of 1x10-12 with an equivalent energy of 80fJ/bit per LED, all without Forward Error Correction (FEC). This achievement highlights the LightBundle platform’s ability to deliver the world’s lowest-power optical interconnects for next generation AI infrastructure by addressing the increasing bandwidth, distance, and reliability requirements most efficiently.

Unlike laser-based optical engines-which cannot operate below a lasing threshold-microLEDs can scale transmit power to nearly arbitrarily low levels, limited primarily by the receiver’s signal-to-noise ratio. While silicon photonics can achieve low effective power by splitting a single external laser across many resonant modulators, microLEDs inherently generate their own light, dramatically simplifying packaging. Avicena microLED transmitters are only a few microns in size, require no temperature stabilization, and avoid complex control loops. These microscopic emitters can be arrayed at extremely high density to deliver terabits of aggregate bandwidth.

This new power-per-lane milestone was enabled by the latest Avicena high-sensitivity receiver technology, developed in collaboration with manufacturing partners and incorporating optimized photodetectors derived from high-volume image-sensor processes.

Instead of serializing low-speed on-chip data (typically around 2 Gb/s per lane) to very high-speed optical lanes, LightBundle transmits raw parallel data directly, dramatically simplifying integrated circuit (IC) input/output (I/O) architectures. This approach enables massive arrays of microLEDs to deliver unprecedented aggregate bandwidth at very low latency, low power consumption, while being easily integrated with any silicon node. LightBundle chiplet transceivers are well-suited to various packaging architectures including co-packaged optics (CPO), on-board optics (OBO), pluggable optical modules, and wide memory interconnects.

“We already demonstrated an efficient microLED link in a live demo at ECOC 2025 in September. By further optimizing our highly sensitive receivers, we have managed to further reduce the operating currents of the microLEDs and obtained Tx energy consumption down further to tens of femtojoules for this part of the link. Combined with the unique properties of microLEDs, we can achieve unmatched energy efficiency in our LightBundle interconnects,” said Rob Kalman, Chief Scientist of Avicena. “This benchmark shows the scalability of our roadmap, how microLED technology can replace legacy laser-based links with a simpler, more reliable and far lower power solution.”

This latest development builds on the ongoing work between Avicena and its hyperscale data center, AI accelerator and memory partners to enable scale-up AI cluster architectures that span multiple racks and thousands of GPUs, dramatically increasing compute performance for emerging agentic AI workloads.

Avicena at SC25:

At Booth 2446, Avicena will demonstrate the critical components of its LightBundle technology platform.

About Avicena

Avicena Tech Corp. is a privately held company located in Sunnyvale, CA, developing LightBundle, a next generation optical interconnect architecture for AI/ML, HPC, sensors, 5G wireless and aerospace applications. This unique, flexible ultra-low energy technology is based on microLEDs, offering both very high bandwidth and low latency. Now, system designers can disaggregate functions like compute and memory and radically grow system throughput. Avicena’s technology is a key building block in the evolution of networking and computing that will reduce the energy impact on our planet.

For more information, visit https://avicena.tech

Fonte: Business Wire

If you liked this article and want to stay up to date with news from InnovationOpenLab.com subscribe to ours Free newsletter.

Related news

Last News

RSA at Cybertech Europe 2024

Alaa Abdul Nabi, Vice President, Sales International at RSA presents the innovations the vendor brings to Cybertech as part of a passwordless vision for…

Italian Security Awards 2024: G11 Media honours the best of Italian cybersecurity

G11 Media's SecurityOpenLab magazine rewards excellence in cybersecurity: the best vendors based on user votes

How Austria is making its AI ecosystem grow

Always keeping an European perspective, Austria has developed a thriving AI ecosystem that now can attract talents and companies from other countries

Sparkle and Telsy test Quantum Key Distribution in practice

Successfully completing a Proof of Concept implementation in Athens, the two Italian companies prove that QKD can be easily implemented also in pre-existing…

Most read

Deepgram Brings Low-Latency Speech Recognition and TTS to Amazon Connect

Deepgram, the world’s most realistic and real-time Voice AI platform, today announced integration of its enterprise-grade speech-to-text (STT) and text-to-speech…

Deepgram Launches Streaming Speech, Text, and Voice Agents on Amazon SageMaker…

Deepgram, the world’s most realistic and real-time Voice AI platform, today announced native integration with Amazon SageMaker AI, delivering streaming,…

MathWorks Showcases AI for Safety-Critical Systems at NeurIPS 2025

NeurIPS 2025, Booth #732 – MathWorks, the leading developer of mathematical computing software, will showcase how engineers and scientists can use MATLAB®…

CORRECTING and REPLACING Block Processes 124 Million Transactions as Americans…

In Lower West Side, Chicago section, first bullet point should read: Total BFCM Weekend Volume: $8.4M (instead of Total BFCM Weekend Volume: $49M). The…

Newsletter signup

Join our mailing list to get weekly updates delivered to your inbox.

Sign me up!