Pioneering Innovations in Quantum Computing, AI and Photonics

Pioneering Innovations in Quantum Computing, AI and Photonics

NTT Research’s Physics and Informatics (PHI) Lab’s mission is to uncover practical, sustainable solutions for real-world challenges by reimagining computation through the lenses of quantum physics, photonic technologies and brain science. By integrating the development of hardware and software, the lab aims to push the boundaries of information processing. Grounded in NTT’s legacy of advancing basic research, the PHI Lab fosters collaboration among physicists, computer scientists, brain scientists and electrical engineers to create a new framework for computation. 

The team at the PHI Lab aims to develop specialized computers and accelerators to enhance modern digital systems for solving a range of complex problems, including combinatorial and continuous optimization, quantum simulation of many-body systems and machine learning. 

The PHI Lab team has attended recent prominent academic conferences, in fields such as photonics, optics, machine learning and leaning representation, to present new research and breakthroughs that are shaping the future of computing and neuro-science:

  • SPIE Optics + Photonics 2024: At SPIE Optics + Photonics 2024, scientists from the PHI Lab and the NTT Tokyo R&D team (Basic Research Lab and Device Technology Lab), made significant contributions through seven presentations, out of which six were invited papers. Specifically from the PHI lab, two invited papers appeared at the conference on research related to overcoming limits to the scalability of programmable photonic circuits and to achieving greater scale for on-chip photonic processors for neural networks.

  • International Conference on Machine Learning (ICML): At ICML 2024, the PHI Lab showcased three key contributions. Visit the ICML website for more details.

  • International Conference on Learning Representation (ICLR): At ICLR 2024, PHI Lab researchers co-authored two key papers: one on “analyzing in-context learning dynamics with random binary sequences, revealing sharp transitions in LLM behaviors” and another on “how fine-tuning affects model capabilities, showing minimal changes.” For more details, visit the ICLR website.

The PHI Lab team also recently published two papers in Optica, a professional society interested in optics and photonics.

  • Mesoscopic ultrafast nonlinear optics: Scientists from the PHI Lab, Stanford and CalTech propose focusing on "quantum optics, in optics" to explore mesoscopic nonlinear optics for advancing quantum information processing.

  • Hyperspectral in-memory computing: PHI Lab researchers propose a hyperspectral in-memory computing architecture using optical frequency combs and programmable memories for energy-efficient matrix-vector operations.

The PHI Lab is committed to continuing to conduct research initiatives focusing on new ways of computation for photonics, advanced algorithms for optimization and learning dynamics. Read more about this research and conferences below. These efforts are also in line with the IOWN (Innovative Optical and Wireless Network) initiatives by pushing the limit further! 

- Kazu

Unlocking the Future of Energy-Efficient Computing

Optical computing has the potential to introduce new ways of computations specifically for AI and machine learning. The advantage of using optics is it requires much lower energy while maintaining (or exceeding) the processing speed. Electronic circuitry has started to face issues, namely its speed and explosive power requirements. Optical computing, which has been described as a “dream” for many years, now has the potential to become reality. Learn more here from Gomi-san’s recent Forbes Tech Council article. (Share on X & LinkedIn).

NTT PHI Lab Advances Programmable Photonics

At SPIE Optics + Photonics 2024 (August 18-22 in San Diego), NTT Research PHI Lab scientists presented two papers on programmable photonics highlighting research related to overcoming limits to the scalability of programmable photonic circuits and to achieving greater scale for on-chip photonic processors for neural networks. The second paper discussed the optimization of multiport interferometers for error robustness and efficient phase shift usage. Read the full blog here. (Share on X & LinkedIn).

Is The Quantum Multiverse Closer Than We Think?

The idea that we could live in a quantum multiverse without necessarily noticing its oddness has been explored by researchers at the Autonomous University of Barcelona in Spain. Their simulation suggests that even in a multiverse where the microscopic world is quantum, the emergence of a non-quantum world may be highly likely. Jess Riedel senior scientist at NTT Research shares that while the findings are a useful addition to studying the fundamentals of quantum theory, further complexity in the model may be needed to definitively reflect our world. Read the full article in New Scientist here.  (Share on X & LinkedIn).

Exploring the Future of AI with Optical Computing

Brian Buntz of R&D World had the opportunity to sit down with Gomi-san to learn more about photonics and optical computing. Brian shared with his audience how NTT Research is focused on tackling long-term challenges in the field of photonics, with the goal of significantly upgrading the efficiency and capabilities of AI. The current energy crisis in AI, driven by the exponential growth in data consumption, has underscored the need for more energy-efficient computing technologies. Read Brian’s article here. (Share on X & LinkedIn).

NTT Unveils New Code Injection Variant at Black Hat

Black Hat brings together ethical white hat defenders and cybersecurity practitioners to really dive deep into the technical aspects of protection suites and architectures. Will Townsend of Forbes was able to attend the conference and hear about NTT Security Holdings' insights on code injection. It's such a crucial topic in the cybersecurity world, and NTT's research identifies a new variant of malicious code injection and offers ways to defend against it. Read Will’s Forbes article here. (Share on X & LinkedIn).

NTT Launches High-Performance International IOWN APN

NTT and Chunghwa Telecom have successfully activated the International IOWN APN between Taiwan and Japan. The APN has demonstrated low latency, no jitter and stable communication with approximately 17 msec on an approximately 3,000 km one-way network. This will be beneficial for various industries and applications, including low-latency interactive immersive entertainment and disaster countermeasures for manufacturing industries. Read the full press release here. (Share on X & LinkedIn).

Press Releases

Blog Posts 

NTT Research In The News

NTT Family Company News

Interested in recent news and updates from NTT family companies? Check out these resources:

Interested in sharing future NTT Research news and articles? Be sure to follow NTT Research on Twitter and LinkedIn to like, retweet and share press releases, blog posts, news and company updates

To view or add a comment, sign in

Explore topics