The race for high-performance computing: optical processors and quantum photonics take central role

In the fast-evolving world of high-performance computing, the explosive growth of AI has sparked a revolution—and a crisis. The insatiable demand for computational power is driving up energy consumption at an alarming rate, creating a bottleneck for further innovation. Enter two groundbreaking technologies based on photonics: optical processors and quantum photonics, each vying to redefine the future of computing, though their paths and challenges differ strongly.

Today, Yole Group once again invites you to gain a deeper understanding of AI and its transformative impact on the high-performance computing ecosystem. Exploring the fusion of photonics and processing could offer a fresh perspective and drive innovative solutions to support this ongoing revolution. Join Eric Mounier, PhD, Principal Analyst at Yole Group, as he analyzes the current state of each technology and demonstrates how they synergize in real time. Grab your coffee and step into the world of high-performance computing!

This analysis is based on a dedicated collection of reports including Neuromorphic Computing Memory and Sensing 2024 – Optical Computing 2024 –Optical Transceivers for Datacom and Telecom 2024 – Silicon Photonics 2024 – Focus on SOI, SiN, and LNOI platforms … More

Optical processors, which use photons instead of electrons, are emerging as a promising solution for computing tasks. These systems boast ultrawide communication bandwidth, ultrahigh processing speeds, and low energy consumption, making them ideal for the matrix multiplication operations at the heart of AI workloads. By leveraging multiplexing techniques like wavelength and spatial mode, optical processors can handle multi-threaded processing with almost no additional overhead. This offers a significant advantage for powering AI inference, simulations, and even safety-critical applications like avionics.

Yet, the road to widespread adoption is far from smooth. Building practical optical logic gates that can rival electronic ones remains a daunting challenge, with issues like cascadability, scalability, and optical loss recovery still unresolved. While research has made strides with simple circuits, large-scale optical computers are still in their infancy. Integration is another hurdle, though advancements in materials like SOI, SiN, TFLN, Graphene, BTO, and polymers are fueling progress in photonic integrated circuits (PICs). Optical processors come in two flavors—analog, driven by electro-optical effects, and digital, where light controls light, the latter being more complex but potentially transformative.

Market projections are optimistic, with the first shipments of optical processors expected around 2027–2028. By 2034, Yole Group estimates nearly 1 million units will be in use, reflecting a staggering 101% compound annual growth rate (CAGR) from 2027 to 2034, with a market value of $2.7 billion (69% CAGR). Prices, initially higher than AI ASICs, are expected to drop as adoption grows among early adopters, OEMs, and systems integrators.

However, not all is rosy in the optical computing space. Several high-profile companies, includingg Celestial AI, Lightmatter, and others, have raised hundreds of millions of dollars to develop optical computing solutions (together, Celestial AI, Lightelligence, and Lightmatter raised about $1.47 billion). However, in a surprising shift, these firms are now changing their focus from optical computing to short-reach interconnection technologies, aiming to address immediate market needs for faster data transfer in data centers. This strategic shift underscores the challenges of bringing optical computing to market and highlights the pressure to deliver tangible results amid high expectations.

On the quantum frontier, quantum photonics is carving out its own niche, using photons to create and manipulate quantum bits, or qubits, for quantum computing. Photon qubits, often called “flying qubits,” travel through optical devices like Mach-Zehnder interferometers to perform quantum gate operations. They come in two forms—discrete variable and continuous variable photons—while other approaches use photons to control non-photon qubits, such as trapped ions or neutral atoms. Companies like IonQ, Quandela, Xanadu, and QuiX are leading the charge, with some, like QCi, exploring optical quantum effects without relying on qubits.

Photon qubits offer distinct advantages: their weak interactions with the environment enable room-temperature operation, long qubit lifetimes, fast gate operations, and high single-qubit gate fidelity. Scalability is being tackled through silicon photonics, but success depends on breakthroughs in components like single-photon sources, detectors, waveguides, and modulators. The quantum market, including quantum photonics, is poised for explosive growth, from $954 million in 2024 to $17.4 billion by 2035, with a 32% CAGR. Post-2028, quantum computing is expected to dominate, and photonics could claim a significant share.

While both optical processors and quantum photonics aim to revolutionize computing, their applications and timelines diverge. Optical processors are geared toward classical tasks like AI inference and simulations, with market readiness on the horizon. Quantum photonics, however, targets quantum applications—optimization, simulation, and AI/machine learning—offering exponential computational power but on a longer timeline. Both face integration and manufacturing hurdles, and without significant use cases, funding, or technological breakthroughs, they risk stalling in the so-called “valley of death.”

Follow Yole Group on its corporate website & LinkedIn

Eric Mounier, PhD, is Chief Analyst, Photonics & Sensing at Yole Group.

With more than 30 years’ experience within the semiconductor industry, Eric provides daily in-depth insights into emerging semiconductor technologies such as quantum technologies, the Metaverse, terahertz, photonics, and sensing.

Based on relevant methodological expertise and a significant technological background, Eric works closely with all of Yole Group’s teams to highlight disruptive technologies and analyze business opportunities through technology & market reports and custom consulting projects.

Eric has spoken at numerous international conferences, presenting Yole Group’s vision of emerging semiconductor technologies, markets, and applications.

Previously, Eric held R&D and Marketing positions at CEA-Leti (France).

Eric Mounier has a PhD in Semiconductor Engineering and a degree in Optoelectronics from the National Polytechnic Institute of Grenoble (France).

Source: www.yolegroup.com