📰 Key Summary

NTT has set up a dedicated fund to bring optical communication technology to AI and picked its first investment target: a Spanish startup. The fund is focused on commercializing the optical communication tech developed under NTT’s IOWN initiative and applying it to AI infrastructure. Notably, NTT isn’t the only backer in this round — Nvidia also joined the funding for the Spanish startup iPronics, a maker of optical switches. The move shows NTT is looking outward for partners to speed up commercialization of its optical tech, positioning it as key infrastructure to support growing AI compute demand. Since the original summary is limited in length, details like the size of this investment round, iPronics’ technical specifics, and NTT-IOWN fund’s follow-on investment plans aren’t provided — check the source link for more.

💬 JudyAI Lab Take

NTT set up a dedicated fund to push optical communication tech into AI infrastructure and picked Spanish startup iPronics as its first investment — and what makes this deal stand out is that Nvidia joined the same funding round.

When major players start treating “optical communication” as a critical piece of AI infrastructure worth investing in, it signals that compute demand growth is bumping up against the physical limits of power and data transfer. Stacking more chips isn’t enough anymore — the actual “wiring” connecting chip to chip and rack to rack has become contested ground. NTT didn’t go it alone on R&D; it went looking for outside partners to speed up commercialization, and Nvidia’s involvement shows this kind of infrastructure bet isn’t a single company’s gamble — it’s a direction the whole AI supply chain is betting on together. For AI builders, this is a reminder that competition at the infrastructure layer is just as fierce as the fight over models and applications.

Worth watching once more technical details on iPronics come out: how optical switches actually end up affecting latency and power consumption in AI clusters.

📅 Source Info

🔗 Further Reading