Nippon Telegraph and Telephone is a company not widely known in the United States, but I believe that will change soon. Based in Japan, NTT employs more than 300,000 people globally, including 2,000 in its research arm in Silicon Valley. That subsidiary, NTT Research, is leaning into a budget of $3 billion annually to conduct scientific investigations related to optical networking, cybersecurity, AI, sustainability, logistics and manufacturing, healthcare and more.
I recently spent time with executives at NTT Research’s Upgrade event in San Francisco, and I came away with a better understanding of the potential impacts of the company’s efforts as they relate to networking and security. I will dive deeper into each of these areas to highlight what I find most noteworthy.
All-Photonics Network
NTT Research’s All-Photonics Network investigation represents a strategic research focus area for the company. It is decidedly a different approach to optical networking, given that it integrates optics-based technology into every network element, including terminals. Typically, silicon photonics research is directed exclusively at the interconnect level within the transport networks that serve as the backbone for connectivity services. By contrast, NTT’s vision is to bring optical efficiency to the entire network infrastructure layer.
Creating an APN is a big effort that will require a significantly rearchitected approach to how networks are constructed, but the benefits could be tremendous. First, latency could theoretically be eliminated through a vertical integration of optical technology. That would support a host of use cases for the application of real-time analytics and automation. Second, the power efficiency that could be realized is a potential game changer. Optical components are highly efficient, generate less heat than alternatives and are less prone to failure as a result. This will be even more important in years to come because the applications for next-generation AI functionality, including generative AI, are power-hungry, and NTT’s vision for an APN could address concerns related to energy consumption. This is a significant point to highlight, given that nearly every enterprise today is focused on promoting sustainable operations and carbon footprint offset initiatives.
NTT’s Innovative Optical and Wireless Network APN investigation is roughly halfway towards its goal of commercialization by 2030. From my perspective, if it comes even close to its goals of a 100x improvement in energy efficiency and transmission capacity, it could revolutionize networking.
A Different Approach To Security
NTT Research is also focused on efforts to reimagine security and data protection. The company uses its Security, Privacy and Integrity Protection Platform to capture its investigations and then productize the results. SPIP aims to incorporate advanced privacy technologies with the goal of simplifying data protection. From my perspective, NTT’s timing is spot-on, given the widespread concerns related to protecting the underlying data used to train large language models that power generative AI applications.
SPIP supports two important elements—attribute-based encryption and multi-party computation. ABE facilitates the ability to segregate data so that users can access only what is needed. NTT Research has made numerous contributions to the ABE standard, including publishing software libraries to allow developers to create commercially available solutions.
Meanwhile, MPC is not a new cryptography concept; its foundation can be traced back to the 1970s. It is designed to allow joint functional computation while maintaining the privacy of inputs through encryption. This is an important consideration, given the challenge of data leakage that results from encryption and decryption schemes. NTT calls its secure computational system San-Shi, and it consists of multiple servers and a client that registers data through a secret sharing process. That might sound ambiguous on the surface, but in dividing the data among multiple servers, no single server can obtain a complete view of the information. The other benefit to NTT’s approach is that there is apparently no tax to performance, as is often associated with encryption; the company says its research indicates that San-Shi’s efficiency is equal to the efficiency of non-encrypted data processing.
Wrapping Up
As I spend more time with NTT, I continue to be impressed by the company’s capabilities and research efforts. The need for more robust forms of networking and security will continue as next-generation applications place pressure on the limits of existing technology infrastructure. NTT’s APN represents the company’s “moonshot,” with SPIP having more immediate impacts today. However, the technology industry needs audacious goals, because these are what often lead to innovation breakthroughs. NTT may be one of the best kept secrets outside of Japan today, but given the company’s deep investments in groundbreaking research, that will likely change soon.
