Latest Trends in Optical Technologies

Using light as a neural network, as this viral video depicts, is actually closer than you think. In 5-10yrs, we could have matrix multiplications in constant time O(1) with 95% less energy. This is the next era of Moore's Law. Let's talk about Silicon Photonics... The core concept: Replace electrical signals with photons. While current processors push electrons through metal pathways, photonic systems use light beams, operating at fundamentally higher speeds (electronic signals in copper are 3x slower) with minimal heat generation. It's way faster. While traditional chips operate at 3-5 GHz, photonic devices can achieve >100 GHz switching speeds. Current interconnects max out at ~100 Gb/s. Photonic links have demonstrated 2+ Tb/s on a single channel. A single optical path can carry 64+ signals. It's way more energy efficient. Current chip-to-chip communication costs ~1-10pJ/bit. Photonic interconnects demonstrate 0.01-0.1pJ/bit. For data centers processing exabytes, this 200x improvement means the difference between megawatt and kilowatt power requirements. The AI acceleration potential is revolutionary. Matrix operations, fundamental to deep learning, become near-instantaneous: Traditional chips: O(n²) operations. Photonic chips: O(1) - parallel processing through optical interference. 1000×1000 matmuls in picoseconds. Where are we today? Real products are shipping: — Intel's 400G transceivers use silicon photonics. — Ayar Labs demonstrates 2Tb/s chip-to-chip links with AMD EPYC processors. Performance scales with wavelength count, not just frequency like traditional electronics. The manufacturing challenges are immense. — Current yield is ~30%. Silicon's terrible at emitting light and bonding III-V materials to it lowers yield — Temp control is a barrier. A 1°C change shifts frequencies by ~10GHz. — Cost/device is $1000s To reach mass production we need: 90%+ yield rates, sub-$100 per device costs, automated testing solutions, and reliable packaging techniques. Current packaging alone can cost more than the chip itself. We're 5+ years from hitting these targets. Companies to watch: ASML (manufacturing), Intel (data center), Lightmatter (AI), Ayar Labs (chip interconnects). The technology requires major investment, but the potential returns are enormous as we hit traditional electronics' physical limits.

Light-Based Quantum Leap: New Protocol Entangles Photons Without Measurement A Groundbreaking Step Toward Scalable Quantum Computing with Light In a major development for quantum physics and photonics, Georgia Tech researchers have proposed a new method to generate entanglement between photons—without relying on quantum measurement. This novel approach may overcome one of the key obstacles in building quantum computers that use light, opening the door to more scalable, reliable, and efficient quantum systems. How the New Protocol Works • The Core Innovation • Traditional methods for entangling photons rely on quantum measurements, which are probabilistic and often inefficient. • Georgia Tech’s protocol instead uses a geometric concept called non-Abelian quantum holonomy, allowing deterministic and consistent entanglement without measurements. • Why This Matters for Photonic Quantum Computing • Photons are ideal carriers of quantum information—they’re fast, stable, and immune to many forms of noise. • However, photons do not naturally interact with one another, making entanglement difficult. • This new method creates interaction-like behavior without requiring the photons to touch or interfere directly. • What Holonomy Enables • Holonomy is a geometric phase acquired by a quantum system over a closed path in its parameter space. • By leveraging non-Abelian holonomies, which depend on the order of operations, researchers can precisely control and entangle photon states. Key People and Publication • Led by Professor Chandra Raman from Georgia Tech’s School of Physics. • Postdoctoral researcher Aniruddha Bhattacharya emphasized the difficulty of photon interaction and the significance of overcoming it. • Findings were peer-reviewed and published in Physical Review Letters, a leading physics journal. Why This Discovery Is Important This innovation marks a critical advance in the pursuit of photonic quantum computers—systems that use light instead of matter-based qubits. Because photonic systems are naturally suited to long-distance communication and faster processing, this breakthrough could significantly accelerate the development of distributed quantum networks, secure communication channels, and large-scale quantum processors. By eliminating the need for quantum measurement during entanglement, the Georgia Tech protocol improves both the efficiency and reliability of quantum operations. This is a vital step toward realizing practical quantum computers that harness the speed and elegance of light. As research in quantum holonomy and photonic systems continues, this work lays foundational principles for a new generation of quantum technologies.

The lights just got a lot brighter at the edge of the #AI universe, and not because of another GPU drop or a new “foundational” model trying to sell you self-awareness. No, this time it’s real tech, actual innovation you can touch, measure, and wire into the future. Avicena Tech just closed a $65 million Series B, led by Tiger Global, with SK hynix, Cerberus Capital Management, Lam Research, Maverick Silicon, and Prosperity7 Ventures backing the play. That puts the total raise at $120 million, which is the kind of runway you don’t get unless you’re building something the #hyperscalers are already losing sleep over. Founded in 2019 by #photonics heavyweight Bardia Pezeshki and CTO Rob Kalman, who both cut their teeth in the wild west of #opticalnetworking, Avicena is flipping #photons into payloads. Their LightBundle™ platform is built on #GaN #microLEDs and integrated #CMOS drivers, delivering over 1Tbps/mm bandwidth density at under 0.5pJ/bit. That’s not evolution, that’s teleportation compared to what’s powering most AI clusters today. This isn’t some academic flex on power curves. This is commercial traction with teeth: Avicena’s 12mm² chiplet clocks in at 1Tbps bidirectional, already winning designs from two top-five cloud players for 2026 deployment. They’ve got SK hynix tapped in for memory-to-processor interconnects. Corning is in with a 331-core fiber that lights up 10 meters of bandwidth without blowing your latency budget. DSPs? Gone. Latency? Sub-5ns. That’s the difference between real-time AI and real slow #inference. And they’re scaling with purpose. TSMC is optimizing #photodetector arrays. ams OSRAM is lined up to pump out a million-plus units per month by 2026. Avicena isn’t just building optical I/O, they’re making it manufacturable. That’s how you go from #deeptech to #datacenter dominance. No science fair here, just science fiction made viable. They’ve also built a team that knows what exits look like. Greg Dougherty (ex-Oclaro CEO) is on the board. Chris Pfistner came over from Inphi / Marvell. Sama Pourmojib is leading ops. The headcount tripled since Series A with hires out of Intel, TSMC, and Micron. That’s not a startup, that’s a vanguard. #AIsystems are hitting the wall on power and bandwidth, and LightBundle is Avicena’s way of punching through. It’s not about plugging the gaps, it’s about redefining the fabric the system runs on. You want #zettascale? You’d better call someone who speaks in picoseconds. This isn’t just interconnect. This is insight in motion. Congrats to the whole crew, Bardia Pezeshki, Rob Kalman, and the Avicena team, for proving that when you combine legacy chops with new-world optics, you don’t just move data. You move markets. #Startups #StartupFunding #VentureCapital #SeriesB #Data #AI #MicroLED #EdgeComputing #DeepTech #Optics #Technology #Innovation #TechEcosystem #StartupEcosystem Thank you, Vention. Without your support, none of this would be possible.

Researchers from Columbia University and Cornell University recently reported a 3D-photonic transceiver that features 80 channels on a single chip and consumes only 120fJ/bit from its electro-optic front ends. The #transceiver achieves low energy consumption through low-capacitance 3D connections between photonics and co-designed #CMOS electronics. Each channel has a relatively low data rate of 10Gbps, allowing the transceiver's electronics to operate with high sensitivity and minimal energy consumption. The large array of channels compensates for the low per-channel data rates, delivering a high aggregate data rate of 800Gbps in a compact transceiver area of only 0.15mm2 (@5.3Tbps/mm2). In addition, having many low-data-rate channels relaxes signal processing and time multiplexing of data streams native to the processor. Furthermore, wavelength-division-multiplexing (#WDM) sources for numerous data streams are becoming available with the advent of chip-scale microcombs. The EIC is bonded to the PIC based on a 15μm spacing and a 10μm bump diameter (@25μm pitch) in an array of 2,304 bonds. This process mitigates two potential failure risks: 1) excessive tin causing flow and electrical short to adjacent bonds and 2) insufficient tin leading to brittle bonds. 👇Figure 1: a) An illustration of the 3D-integrated photonic-electronic system combining arrays of electronic cells with arrays of photonic devices. b) A microscope image of the 80-channel photonic device arrays with an inset of two transmitter and two receiver cells. c) Microscope images of the photonic and electronic chips. The active photonic circuits occupy an area outlined in white, while the outer photonic chip area is used to fan out the optical/electrical lanes for fiber coupling and wire bonding. The blue overlay shows a four-channel transmitter and receiver #waveguide path; the disk and ring overlays are not to scale. An inset shows a diagram of the fiber-to-chip edge coupler, consisting of a silicon nitride (Si3N4) inverse taper and escalator to silicon. d) A scanning electron microscope image of the bonded electronic and photonic chip cross-section. e) An image of the wire-bonded transceiver die bonded to a printed circuit board and optically coupled to a fiber array with a US dime for scale. f) A cross-sectional diagram of the electronic and photonic chips and their associated material stacks. Both chips consist of a crystalline silicon substrate, doped-silicon devices and metal interconnection layers. Daudlin, S. et al. Three-dimensional photonic integration for ultra-low-energy, high-bandwidth interchip data links. Nat. Photon. (2025).👉https://lnkd.in/gpeVGZna #SemiconductorIndustry #Semiconductor #Semiconductors #AI #HPC #Datacenter #Optics #Photonics #SiliconPhotonics #Optical #Networking #OCI #Ethernet #Infrastructure #Interconnect #CloudAI #AICluster AIM Photonics TSMC Defense Advanced Research Projects Agency (DARPA) #FiberCoupling #SiP

𝗙𝗿𝗼𝗺 𝗕𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸 𝘁𝗼 𝗕𝗿𝗲𝗮𝗸𝘁𝗵𝗿𝗼𝘂𝗴𝗵: 𝗛𝗼𝘄 𝗣𝗵𝗼𝘁𝗼𝗻𝗶𝗰𝘀 𝗜𝗻𝘁𝗲𝗿𝗰𝗼𝗻𝗻𝗲𝗰𝘁𝘀 𝗮𝗿𝗲 𝗥𝗲𝘄𝗶𝗿𝗶𝗻𝗴 𝘁𝗵𝗲 𝗙𝘂𝘁𝘂𝗿𝗲 𝗼𝗳 𝗔𝗜 & 𝗗𝗮𝘁𝗮 𝗖𝗲𝗻𝘁𝗲𝗿𝘀 The future of AI and high-performance computing won’t be defined by silicon alone. 𝗔𝘀 𝗺𝗼𝗱𝗲𝗹𝘀 𝘀𝗰𝗮𝗹𝗲, 𝗺𝗼𝘃𝗶𝗻𝗴 𝗱𝗮𝘁𝗮—𝗻𝗼𝘁 𝗷𝘂𝘀𝘁 𝗰𝗼𝗺𝗽𝘂𝘁𝗶𝗻𝗴—𝗵𝗮𝘀 𝗯𝗲𝗰𝗼𝗺𝗲 𝘁𝗵𝗲 𝗿𝗲𝗮𝗹 𝗯𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸 𝗳𝗼𝗿 𝗮𝗰𝗰𝗲𝗹𝗲𝗿𝗮𝘁𝗼𝗿𝘀. The limits of copper wires are now holding back bandwidth, power efficiency, and ultimately, AI’s progress. 𝗖𝘂𝗿𝗿𝗲𝗻𝘁 𝗖𝗵𝗮𝗹𝗹𝗲𝗻𝗴𝗲𝘀  • 𝗘𝘀𝗰𝗮𝗹𝗮𝘁𝗶𝗻𝗴 𝗽𝗼𝘄𝗲𝗿 𝘂𝘀𝗮𝗴𝗲: High-speed electrical I/O burns enormous power, especially as bandwidth demands rise.  • 𝗕𝗮𝗻𝗱𝘄𝗶𝗱𝘁𝗵 𝗯𝗼𝘁𝘁𝗹𝗲𝗻𝗲𝗰𝗸𝘀: Copper wires face a ceiling for how much data they can carry, with signal degradation and crosstalk worsening at higher speeds.  • 𝗟𝗮𝘁𝗲𝗻𝗰𝘆 & 𝘀𝗰𝗮𝗹𝗶𝗻𝗴: Traditional interconnects add latency, and scaling to larger multi-chip or multi-rack systems often requires even more energy and complex routing. 𝗣𝗵𝗼𝘁𝗼𝗻𝗶𝗰𝘀: 𝗧𝗵𝗲 𝗦𝗼𝗹𝘂𝘁𝗶𝗼𝗻 #Photonics - using light instead of electricity to move data—offers a path to break through these barriers:  • 𝗨𝗹𝘁𝗿𝗮-𝗵𝗶𝗴𝗵 𝗯𝗮𝗻𝗱𝘄𝗶𝗱𝘁𝗵: Photonic links deliver terabits per second between chips, boards, and racks.  • 𝗟𝗼𝘄𝗲𝗿 𝗽𝗼𝘄𝗲𝗿 𝗽𝗲𝗿 𝗯𝗶𝘁: Photonics reduces wasted energy as heat, enabling higher density and sustainability.  • 𝗟𝗼𝗻𝗴𝗲𝗿 𝗿𝗲𝗮𝗰𝗵, 𝗹𝗼𝘄𝗲𝗿 𝗹𝗮𝘁𝗲𝗻𝗰𝘆: Optical signals maintain integrity over longer distances, crucial for modular and disaggregated architectures. 𝗞𝗲𝘆 𝗛𝘂𝗿𝗱𝗹𝗲𝘀 𝗳𝗼𝗿 𝗠𝗮𝗶𝗻𝘀𝘁𝗿𝗲𝗮𝗺 𝗔𝗱𝗼𝗽𝘁𝗶𝗼𝗻  • 𝗖𝗠𝗢𝗦 𝗶𝗻𝘁𝗲𝗴𝗿𝗮𝘁𝗶𝗼𝗻: Integrating lasers, modulators, and photodetectors with silicon is still complex.  • 𝗣𝗮𝗰𝗸𝗮𝗴𝗶𝗻𝗴 & 𝘆𝗶𝗲𝗹𝗱: High-precision assembly is required; small misalignments can hurt performance and scale-up.  • 𝗧𝗵𝗲𝗿𝗺𝗮𝗹 𝗺𝗮𝗻𝗮𝗴𝗲𝗺𝗲𝗻𝘁: On-chip lasers and drivers add new thermal challenges.  • 𝗖𝗼𝘀𝘁 & 𝗲𝗰𝗼𝘀𝘆𝘀𝘁𝗲𝗺: Photonic components are costlier so volume manufacturing and mature standards are just emerging.  • 𝗦𝗼𝗳𝘁𝘄𝗮𝗿𝗲/𝗮𝗿𝗰𝗵𝗶𝘁𝗲𝗰𝘁𝘂𝗿𝗲: Fully exploiting photonics requires new networking stacks, protocols, and sometimes rethinking system design. 𝗣𝗵𝗼𝘁𝗼𝗻𝗶𝗰𝘀 𝗶𝘀 𝗻𝗼 𝗹𝗼𝗻𝗴𝗲𝗿 𝗷𝘂𝘀𝘁 𝗮 𝗿𝗲𝘀𝗲𝗮𝗿𝗰𝗵 𝘁𝗼𝗽𝗶𝗰—𝗶𝘁’𝘀 𝗻𝗼𝘄 𝘂𝗻𝗹𝗼𝗰𝗸𝗶𝗻𝗴 𝗻𝗲𝘄 𝗳𝗿𝗼𝗻𝘁𝗶𝗲𝗿𝘀 𝗶𝗻 𝗽𝗲𝗿𝗳𝗼𝗿𝗺𝗮𝗻𝗰𝗲 𝗮𝗻𝗱 𝗲𝗳𝗳𝗶𝗰𝗶𝗲𝗻𝗰𝘆 𝗳𝗼𝗿 #𝗔𝗜 𝗮𝗻𝗱 𝗰𝗹𝗼𝘂𝗱 #𝗰𝗼𝗺𝗽𝘂𝘁𝗶𝗻𝗴. The transition from electrons to photons is happening, but its tipping point will depend on integration, ecosystem, and system design breakthroughs. 𝗪𝗵𝗲𝗿𝗲 𝗱𝗼 𝘆𝗼𝘂 𝘀𝗲𝗲 𝘁𝗵𝗲 𝗯𝗶𝗴𝗴𝗲𝘀𝘁 𝗵𝘂𝗿𝗱𝗹𝗲𝘀—𝗼𝗿 𝗼𝗽𝗽𝗼𝗿𝘁𝘂𝗻𝗶𝘁𝗶𝗲𝘀—𝗳𝗼𝗿 𝗽𝗵𝗼𝘁𝗼𝗻𝗶𝗰𝘀 𝗶𝗻 𝗿𝗲𝘀𝗵𝗮𝗽𝗶𝗻𝗴 𝗱𝗮𝘁𝗮 𝗺𝗼𝘃𝗲𝗺𝗲𝗻𝘁 𝗮𝘁 𝘀𝗰𝗮𝗹𝗲? Hrishi Sathwane Tarun Verma Harish Wadhwa Dr. Satya Gupta

Integrated photonic optical switches have recently gained increased attention due to the growing interest in AI and the associated challenges of #AI data communication in #datacenter and #HPC systems. Our group has been active in this area, focusing on the development of novel #photonic switching elements, improved switch #topologies, and fast, scalable, performance-aware #reconfiguration solutions. To provide a comprehensive overview of the state-of-the-art, as well as highlight existing challenges and #research opportunities, we recently published an in-depth #survey on integrated photonic switch fabrics. The review covers photonic switching devices (including silicon photonics, III-V, MEMS, etc.), switch fabric topologies, tuning and reconfiguration mechanisms, and control planes and strategies for managing switch fabric operation. Kudos to my talented Ph.D. student, M.Amin Mahdian for his dedicated effort in putting together this comprehensive and valuable survey paper. Also, special thanks to #NSF for supporting this work. Enjoy! Link to the paper (open access): https://lnkd.in/gbi_dMhY #siliconphotonics #switch #network #interconnect #photonics #control #communication

The future of AI isn't just about smarter algorithms, it's about reimagining the physical infrastructure that powers them While the industry focuses on the latest AI models, a fundamental shift is happening at the hardware level POET Technologies just announced a $25M strategic investment to accelerate their breakthrough in optical computing for AI systems Here's why this matters: Traditional AI systems are increasingly bottlenecked by the physics of moving data through copper interconnects POET's revolutionary POET Optical Interposer™ technology changes the game by enabling seamless integration of electronic and photonic devices into a single chip, solving bandwidth and latency problems that become critical as AI workloads scale exponentially The impact is significant: • Bandwidth: Optical interconnects carry vastly more data than electrical connections • Power efficiency: Photonic systems consume significantly less power for data transmission • Speed: Light-based communication eliminates delays inherent in electronic systems This isn't just incremental improvement, it's infrastructure transformation From chip-to-chip communication within AI servers to high-speed connections supporting 800G and 1.6T+ data rates, optical computing addresses the physical limitations that could otherwise constrain AI advancement The market validation is compelling too POET has raised over $100M in equity capital at increasingly higher prices over the past year, with strong interest from institutional and strategic investors recognizing the compelling value proposition for AI networks and systems As AI infrastructure spending explodes and organizations race to build more capable systems, the companies that recognize and invest in optical computing early will likely gain sustainable competitive advantages The physics of light create natural performance moats that can't be replicated through software optimization alone The next generation of AI won't just think differently, it will be built differently Read more: https://lnkd.in/gUzFK3RM #AI #OpticaIComputing #AIInfrastructure #TechInnovation #FutureOfAI #DataCenters #QuantumComputing #TechLeadership

According to fibeReality’s latest intelligence, there is a rather shocking level of hopefulness at two of the most prominent players on the demand side, involving the future use of silicon photonics modulators at 400G per lane, without the requirement for hybrid materials (please see: https://lnkd.in/euvZjRdT). One of them is Google (please also see: https://lnkd.in/e2A95QdV). It is involved with a few multi-project wafer runs, and the expectation is that the cloud services provider will be publishing data soon, which will presumably show positive results. Google is usually the leader in influencing optics technology direction, with rare exceptions, such as with its optical circuit switches (please see: https://lnkd.in/gnDAeTay). In fact, its use of the OCS may be the principal reason for needing to move in this direction, as the switch may not be maintaining the polarization extinction ratio well, along with explaining why the large user has been building its own SiPh PICs. Yet, while the necessity for moving in this direction could be confined, it could open the door to quicker development that will be attractive to the entire market. (Otherwise, the principal expectation in the marketplace continues to heavily lean towards SiPh running out of gas at 200G, and that pulling off such a solution at the higher rate within four years is likely unreasonable.) Most significantly, despite SiPh being awfully me-too, and remaining in the emerging stage of its product life cycle for a quarter of a century, the material platform has become an incumbent solution. It is another demonstration that as with InP EMLs, the optics space up and down the food chain remains conservative in nature, and hesitant to move in a revolutionary direction, unless mandatory.

Considering how quickly telecommunications has advanced, we can see that 6G is quickly approaching reality. With the help of tech giants like Google, Microsoft, Intel, and Sony, the IOWN Global Forum is leading this change in all-photonic networks #APNs. These networks have the potential to completely change our digital infrastructure since they transfer data using light instead of electrical impulses. Photonic networks are set to dramatically reduce energy consumption and enhance data handling capabilities. This is particularly important as the demand for AI and digital services continues to grow. The impact is far-reaching--- improved efficiency, sustainability, and performance. In the financial sector, photonic technology offers greater resilience and efficiency. Banks can operate data centers across multiple locations, ensuring seamless and secure transactions. This is crucial for maintaining the integrity and speed of financial operations. Institutions like Mitsubishi UFJ are already exploring these technologies to enhance their infrastructure. The media industry also stands to benefit significantly. With increasing demand for streaming services, platforms like Netflix and Amazon Prime require vast amounts of data to deliver content. Photonic networks can make this process more energy-efficient and capable of handling higher data volumes with lower latency. Sony's work with NTT to develop a wide-area remote production platform highlights the potential of this technology in improving broadcast and media streaming capabilities. Moreover, deploying 6G antennas will be more cost-effective and efficient with photonic networks. By using optical fiber to connect antennas to radio data centers, telecom carriers can optimize network performance and share radio towers, reducing the heavy investments required for antenna deployments. This will facilitate a smoother and more economical transition to 6G. Looking ahead, it's clear that all-photonic networks will play a crucial role in shaping the next generation of mobile networks. These advancements are building a more sustainable and efficient digital future. #TechInnovation #6G #Telecommunications #Innovation #Sustainability #AI #Tech #Telecom

While this isn’t breaking news, Ayar Labs' $155 Million Series D announced last week is the capstone of an amazing year of photonics funding that will help communications infrastructure scale to the next level in this “Age of AI”. 2024 was marked by explosive growth in the size and complexity of large language models, with leaders like Meta, Open AI, Cerebras, SambaNova, and Groq competing for breakthrough performance in delivery speed. But this often comes with tradeoffs: requiring more hardware and greater power consumption to deliver blazing inference. AI infrastructure is at a tipping point, with projections exceeding $1 trillion in investments over the next decade. Such massive scaleout places big burdens on data center operators to deliver at the highest levels, and costs are soaring which COULD jeapordize the availability and spread of AI, if only the companies with large war chests can afford to deploy it. So, the industry is racing to eliminate inefficiencies that throttle performance and increase efficiency, not only as a means of bringing the cost of delivery down, but as a way to help AI proliferate responsibly. Traditional copper interconnects and pluggable optics, which are the longstanding backbones of data movement, are now viewed as bottlenecks in the face of AI’s exponential growth. The solution? A fundamental shift in how data is moved. Enter optical I/O. The potential for optical I/O wasn’t just discovered. Ayar and a few other compelling startups have been working on it for years but Ayar’s laser-like focus, technology stack and their core talent give them a competitive edge… as well as their strategic approach to gaining investment. Founded in 2015 by renowned experts in the field of photonics, they sought strategic investors that are all purveyors of next-gen hardware companies with high potential to create breakthrough solutions to advance high-performance computing “The backing of industry giants and growth-focused investors underscores the transformative potential of our technology,” Mark Wade, shared in their press release. “Optical I/O is not just a performance improvement—it’s a fundamental enabler of the future AI infrastructure ecosystem.” Ayar’s impressive demonstrations at SC24 coupled with this investment sends a strong signal: They’re ready to deliver and they’ve called their shot for 2025—expand the team, scale manufacturing, and solidify optical I/O as the cornerstone of next-gen AI systems. All of this should be exciting to engineers and leaders who have been deeply-focused on advancing optical technology for their entire careers. Look to Ayar for fresh opportunities to be part of something big because the future is here, and apparently, it’s being built with light. #semiconductorindustry #photonics #artificialintelligence