Analogue-to-digital converters serve as crucial interfaces that translate real-world continuous signals into discrete digital data suitable for computational analysis. Conventional electronic ADCs, while incredibly advanced, face intrinsic bandwidth limits caused by electronic component speeds and power consumption constraints. This impasse stymies progress in applications demanding ultra-broadband digitization, such as radar imaging, high-frequency communications, and ultrafast spectroscopy. To transcend these limitations, the research team turned to photonic technologies, specifically Kerr soliton microcombs, known for their ability to generate stable, precisely spaced optical frequency lines across vast spectral bandwidths.

Kerr soliton microcombs are generated in ultra-high-Q microresonators by exploiting the Kerr nonlinear optical effect. When pumped with a continuous-wave laser, these microresonators produce a coherent train of equally spaced frequency lines—or comb teeth—that can serve as a multi-wavelength carrier source. The particular formation of soliton pulses within these resonators ensures not only spectral purity but also temporal stability critical for high-fidelity signal processing. Integrating these microcombs into ADC architectures opens novel pathways for photonic-assisted sampling mechanisms, breaking the bottleneck of traditional electronics.

At the heart of the reported system is a synergistic design that channels the microwave-frequency input signal through a photonic front-end employing Kerr soliton microcombs to perform optical sampling. By effectively translating electronic signals into the optical domain and mapping them onto different comb lines, the ADC achieves a sampling rate far exceeding what purely electronic devices could muster. Subsequently, photodetectors convert the optically sampled signals back into the electrical domain for digital reconstruction. This electronic-photonic hybrid approach grants the converter an effective bandwidth of 320 GHz, marking a significant leap in sampling frequency and resolution fidelity.

One of the prominent challenges the researchers addressed was maintaining signal integrity amidst the complex interplay of nonlinear optics, laser stabilization, and electronic data processing. The generation and stabilization of the microcomb required precise control of pump laser parameters and resonator temperature to sustain the dissipative Kerr soliton state without drift or disruption. Fine-tuning these variables ensured the generated comb lines remained phase-locked and temporally coherent, which is essential for accurate time-domain sampling and minimizing jitter-induced errors in the analogue-to-digital conversion process.

Beyond the microcomb generation, the team engineered an advanced microwave photonic sampling module featuring dispersion-compensated waveguides and high-speed photodetectors. This enabled efficient modulation of incoming analogue signals across the comb spectrum and preserved the ultra-broadband sampling characteristics. Coupled with low-noise electronic analog front-end circuits and high-speed analog-to-digital converters for the post-photonic stage, the entire system efficiently bridged the optical-electronic divide with minimal added noise or distortion.

The implications of successfully achieving a 320 GHz photonic-electronic ADC extend well beyond laboratory demonstrations. In modern communications, handling radio-frequency signals at these frequencies is essential for emerging 6G and future wireless standards targeting terabit-per-second throughput. The ability to digitize such high-frequency analog signals directly supports advanced modulation schemes, massive MIMO antenna arrays, and real-time spectrum analysis with unprecedented granularity. Moreover, ultrafast ADCs integrated into radar and sensing equipment can facilitate improved resolution, range, and target identification capabilities, benefiting aerospace, defense, and autonomous systems.

Furthermore, the breakthroughs in utilizing Kerr soliton microcombs for analogue-to-digital conversion could spur innovations in quantum information sciences and neuromorphic computing. Leveraging finely spaced comb lines promises to enhance multiplexing density and parallelism, which are vital for scaling quantum communication channels and hardware neural networks. The precise control over pulse timing and spectral characteristics inherent to soliton microcombs can enable deterministic quantum state preparation and measurement, while the high sampling frequencies align with the rapid data throughput demands of artificial intelligence accelerators.

Notably, this research aligns with a growing trend of hybrid photonic-electronic systems that seek to harness the speed of light and the versatility of electronics in tandem. By eschewing purely electronic bottlenecks, researchers can now circumvent the RC time constants and heat dissipation challenges that plague high-frequency electronic circuits. Photonic integration, facilitated by advances in microfabrication and silicon photonics, allows these ADC systems to be miniaturized and potentially scaled for commercial deployment. This opens pathways for compact, energy-efficient, and high-performance digitization modules tailored for edge computing and data center applications.

An additional facet of the study addresses the linearity and dynamic range performance of the 320 GHz ADC system. Analog-to-digital conversion quality is judged not just by sampling rate but also by the integrity with which signal amplitude variations are captured. The researchers optimized the system architecture to minimize intermodulation distortion and spurious noise components, leveraging the intrinsic low-noise features of soliton microcomb generation and high-fidelity photodetection. Such meticulous engineering ensures high effective number of bits (ENOB), providing acceptable quantization error levels for demanding signal processing tasks.

The stability of the system over extended operation periods was another key consideration. Long-term drift in comb line frequencies or pump parameters could degrade conversion accuracy. Implementing active feedback loops and temperature stabilization enabled the prototype to maintain robust performance, illustrating the feasibility of real-world deployment. Future iterations integrating on-chip resonator temperature sensors and feedback electronics promise further enhancements in operational stability and environmental tolerance.

Importantly, this innovative ADC concept underscores the confluence of materials science, nonlinear optics, and microwave engineering in solving complex challenges. The microresonators utilized were fabricated with ultra-smooth surfaces and high-quality materials to minimize optical losses, which directly influence comb generation efficiency and stability. Advancements in these fabrication technologies were instrumental in realizing a compact photonic platform capable of supporting the demanding requirements of high-speed ADC applications.

While the immediate focus of this work is on photonic-electronic analogue-to-digital conversion, the underlying principles may extend to other ultrafast optical signal processing domains. The comb-based sampling technique could be adapted for optical arbitrary waveform generation, frequency synthesizers, and high-precision timing distribution networks. Such versatility may catalyze cross-disciplinary research bridging photonics, electronics, and information theory, driving future innovations in communication and sensing technologies.

This remarkable demonstration of a 320 GHz photonic-electronic ADC exploiting Kerr soliton microcombs represents a transformative stride in signal acquisition and processing technology. By harnessing the power of nonlinear optics and integrating it seamlessly with high-speed electronics, the research not only surmounts longstanding bandwidth obstacles but also paves the way for novel applications demanding ultrahigh sampling rates. As technology trends increasingly favor photonic integration and hybrid systems, this approach sets a new benchmark for performance, inspiring further exploration of soliton microcombs in next-generation information processing platforms.

Looking ahead, the researchers envision continued development focused on improving integration density, reducing system complexity, and exploring scalable manufacturing methods. Efforts to integrate the entire photonic-electronic ADC on chip, incorporating low-loss waveguides, modulators, and detectors, could drastically reduce latency and cost, making the technology viable for widespread use. Additionally, exploring novel resonator materials and designs may enable tuning of comb spectra to even higher frequencies or broader bandwidths, expanding the ADC capabilities further.

In conclusion, the marriage of Kerr soliton microcombs with analogue-to-digital converting technologies emerges as a highly promising route to breaking through the frequency and resolution barriers of current systems. The 320 GHz ADC demonstrated marks a milestone, showcasing how interdisciplinary innovation can unlock functionality critical to future communication, sensing, and computing infrastructures. As this exciting field advances, the ripple effects on scientific instrumentation and industrial applications are poised to be profound and far-reaching.

Subject of Research: Photonic-electronic analogue-to-digital conversion enabled by Kerr soliton microcombs

Article Title: 320 GHz photonic-electronic analogue-to-digital converter (ADC) exploiting Kerr soliton microcombs

Article References:
Fang, D., Drayss, D., Peng, H. et al. 320 GHz photonic-electronic analogue-to-digital converter (ADC) exploiting Kerr soliton microcombs. Light Sci Appl 14, 241 (2025). https://doi.org/10.1038/s41377-025-01778-1

Image Credits: AI Generated

DOI: https://doi.org/10.1038/s41377-025-01778-1

Quantum mechanics inherently defies the deterministic nature of classical computers; they cannot effectively manage the uncertainty and complexity found within quantum systems. In 1981, Richard Feynman posed the significant question regarding whether computers could efficiently simulate such systems. Traditional binary computing systems falter in this realm, as they encode information in a binary format, offering limited functionality in probabilistic scenarios. The vision of quantum computers as a solution remains tantalizing yet faced with numerous technical hurdles. In parallel, researchers are exploring the concept of probabilistic computing, which endeavors to efficiently solve complex problems by embracing uncertainty.

At the heart of this paradigm shift lies the probabilistic bit, or p-bit. Unlike conventional bits that operate strictly in binary states of 0 and 1, p-bits exist in a state of flux, oscillating between these values. This dynamism enables a new approach to computing, one that taps into the inherent randomness found in physical systems, notably through thermal fluctuations. The design of p-bits must balance efficiency and stability, presenting challenges and opportunities for material scientists and engineers alike.

Recent advancements have seen a team from Fudan University, spearheaded by Professor Jian Shen and Hangwen Guo, successfully fabricate p-bit devices using manganite nanowires. This innovative material exploits the phase separation between ferromagnetic and antiferromagnetic states, allowing these devices to transition between low resistance (representing 0) and high resistance (representing 1). This transition is not merely theoretical; it has been demonstrated experimentally through precise control with nanoampere-level currents. This level of control is essential for the stability and reliability needed in practical computational applications, and it brings p-bits a step closer to widespread utilization.

What sets this research apart is not only the successful demonstration of operating p-bits but also the exceptional stability these devices exhibit. During extensive testing, the operational stability of the p-bits has been remarkable, with variations kept within a standard deviation of less than 1.3%. Stability is a critical aspect, especially in computational scenarios that require repeated operations. This finding substantiates the viability of p-bits in real-world applications, whether in optimization problems or complex simulations.

The implications of these p-bits extend far beyond mere theoretical benefits. In practical terms, simulations have showcased their critical role in tasks requiring Bayesian inference—a methodology widely applied in statistics and machine learning. The accuracy of the results derived from these p-bits was found to significantly surpass those yielded by conventional probabilistic bits. This leap in performance has profound implications, positioning this technology favorably against existing solutions while offering a viable path towards high-performance probabilistic computing.

Moreover, the device’s ability to generate high-quality intrinsic true random numbers opens new horizons in cryptographic applications. Randomness plays a pivotal role in secure communications, and harnessing a device capable of producing reliable random numbers is a commendable breakthrough in this field. As digital security threats continue to evolve, innovations like this provide not just solutions but a proactive stance against the risks associated with data usage.

This fusion of classical and quantum principles encapsulated within these manganite nanowires serves as a bridge, intertwining the established frameworks of classical computing with the promising potentials of quantum technologies. The findings from Fudan University offer a glimpse into a future where such hybrid systems could dominate computing. As researchers delve deeper, the continuing exploration of material properties and behaviors is likely to unveil even more pathways toward optimizing probabilistic computing.

The impact of these advancements is reflected not only in academia but also across industries that rely on complex computations and analyses daily. Whether enhancing logistics through optimization models or driving forward artificial intelligence algorithms, the applications of p-bits promise to permeate various sectors. Consequently, the research team’s contributions may herald a new era of computational technology.

Amidst this technological renaissance, it’s crucial to address the ongoing challenges in scaling these technologies for commercial use. While the prospects are promising, engineers and scientists will need to collaborate to overcome existing hurdles such as manufacturing processes, integration with classical systems, and data management. Bridging these gaps will be essential for transitioning theoretical advancements into tangible solutions that can benefit society at large.

In conclusion, the advent of probabilistic computing through the successful implementation of p-bits establishes a pivotal milestone in the evolution of computer science. The findings not only underscore the potential of manganite nanowires in this domain but also provide a roadmap toward realizing robust probabilistic computing systems. As research progresses, it is anticipated that such developments will ignite further innovations, ultimately enhancing our computational capabilities and understanding of the universe.

The journey of computing continues. With each breakthrough, we approach a deeper understanding of the mysteries that intertwine the classical and quantum realms and potentially revolutionize the way we interact with information technology.

Subject of Research: Probabilistic computing using manganite nanowires
Article Title: Superior probabilistic computing using operationally stable probabilistic-bit constructed by manganite nanowire
News Publication Date: 2025
Web References: http://dx.doi.org/10.1093/nsr/nwae338
References: National Science Review
Image Credits: ©Science China Press
Keywords: probabilistic computing, quantum mechanics, p-bits, manganite nanowires, Bayesian inference, cryptography, information technology, stability, optimization, true random numbers, classical computing, quantum computing