EP964: Business at the Speed of Light

Masahisa Kawashima, IOWN Technology Director at NTT, discusses how IOWN (Innovative Optical and Wireless Network) technology will transform financial services. IOWN's photonic infrastructure enables high-speed, low-latency, and highly reliable data transfer, essential for financial institutions requiring seamless AI and data-driven operations. It allows for efficient, scalable, and sustainable data center operations by reducing power consumption through decentralized setups in suburban areas. IOWN’s infrastructure will also enhance security, operational resilience, and disaster recovery. The technology aims to provide banks with high-bandwidth, low-latency connectivity for AI computing, improving business continuity and data management. While large-scale deployment may take a few years, IOWN is positioning itself as a key enabler for the next generation of digital financial services

Transcript

SPEAKER_01 0:00

You know, when you fire off a quick prompt to a generative AI, or you know, you stream a massive 4K movie to your living room, you probably imagine all that data just kind of floating effortlessly up into some pristine and visible cloud.

SPEAKER_00 0:14

Right. It feels entirely weightless to the user.

SPEAKER_01 0:17

Exactly. I mean, you click a button and the magic just happens over the air. But the reality is that our digital world is incredibly heavy. It's very physical. Oh, absolutely. It is grounded in these massive, humming concrete buildings filled with millions of miles of copper wires and giant spinning fans. And right now, that physical infrastructure is sweating.

SPEAKER_00 0:39

Aaron Powell Sweating is probably an understatement at this point.

SPEAKER_01 0:41

Yeah, fair enough. But that's why we're here today. We are about to explore a groundbreaking shift in how the world's digital information travels. We are talking about fundamentally rewriting the physical rules of the Internet.

SPEAKER_00 0:52

Aaron Powell Basically moving from the age of electronics to the age of photonics.

SPEAKER_01 0:56

Aaron Powell Right. Literally doing business at the speed of light.

SPEAKER_00 0:59

Aaron Powell It's a profound transition. I mean, we are moving away from a physical paradigm that has dominated since the dawn of computing. We've pushed our current materials to their absolute physical limits. And now, well, we have to engineer a completely new medium.

SPEAKER_01 1:14

Aaron Powell And to figure out how we actually pull that off, we are diving into some really fascinating research today for this deep dive.

SPEAKER_00 1:20

Aaron Powell We have some incredibly solid sources to work from.

SPEAKER_01 1:23

We do. We've got hard market data from an omnial white paper titled The All Photonics Network Enables the Next Gen Digital Economy.

SPEAKER_00 1:31

Aaron Powell That's a vital piece of the puzzle.

SPEAKER_01 1:32

Aaron Powell Yeah. And we are pairing that with a really eye-opening December 2025 interview from the IBSI FinTech Journal. That interview features Masahisa Koshima, who is the IOWN technology director at NTT.

SPEAKER_00 1:46

Aaron Powell Right, the person actually helping to architect this new physical reality for our networks.

SPEAKER_01 1:50

Aaron Powell Exactly. Okay, let's unpack this because you listening to this right now, you might think our current internet infrastructure is totally fine. I mean, your Wi-Fi connects, your apps load, your video calls mostly work. For the most part, yeah. But behind the scenes, the AI revolution is pushing our power grids and our data networks to the absolute breaking point. We are hitting a very real, very physical wall.

SPEAKER_00 2:13

Aaron Powell We really are. And to truly understand the solution that Kawashima is proposing, we first have to grasp the sheer staggering scale of the physical data crisis. Aaron Powell Right.

SPEAKER_01 2:23

The Omdeo White Paper lays this out in pretty stark terms.

SPEAKER_00 2:27

Aaron Powell It does. They forecast that total global data center capacity is expected to grow by 57% from 2024 to 2027 alone.

SPEAKER_01 2:36

Wow. 57% in just three years.

SPEAKER_00 2:39

Yeah, and crucially, they don't measure this growth in square footage or server racks. They measure it in total power load in gigawatts.

SPEAKER_01 2:45

Aaron Powell Wait, gigawatts? That is terrifying because a single gigawatt is roughly the output of a nuclear reactor, right? Yeah. Or enough to power a mid-sized city.

SPEAKER_00 2:52

Exactly. We are talking about adding the power demands of dozens of cities to the global grid in a span of 36 months.

SPEAKER_01 2:59

And what's driving that massive jump? I mean, it's almost entirely the need to support new AI capabilities, isn't it?

SPEAKER_00 3:05

Oh, without a doubt, the technology ecosystem surrounding artificial intelligence is incredibly power hungry. We're no longer just retrieving static text files or images from a server.

SPEAKER_01 3:15

Right. It's way more complex now.

SPEAKER_00 3:17

Generative AI models are processing unimaginably vast data sets in real time. They're running billions of mathematical calculations the second you hit enter on a prompt.

SPEAKER_01 3:28

I was trying to picture why doing that on our current infrastructure is causing such a massive power draw. And it really comes down to the physics of the materials we use, doesn't it?

SPEAKER_00 3:38

It really does. For the last century, we've relied heavily on copper.

SPEAKER_01 3:41

So it sounds like we are trying to force an entire ocean of AI data through the physical equivalent of a garden hose, and the friction is literally burning us out.

SPEAKER_00 3:51

That is a highly accurate way to visualize the mechanism at play. I mean, Kawashima points this out. Electrons have mass. Right. When you force them through a copper lattice, they collide with atoms. That creates electrical resistance. And in physics, resistance inevitably generates heat.

SPEAKER_01 4:06

So a massive portion of the energy we pump into a data center isn't even actually doing the computing.

SPEAKER_00 4:11

Exactly. It's just overcoming the physical friction of the wire. And then because the servers are generating so much heat from that friction, you have to pump in even more gigawatts of power to run industrial cooling systems.

SPEAKER_01 4:23

Just to stop the whole building from melting down.

SPEAKER_00 4:25

Pretty much.

SPEAKER_01 4:26

I want you, the listener, to just imagine that physical reality for a second. Every time you ask an AI to summarize a long PDF for work, you are relying on these massive, power-hungry buildings that are literally fighting against the laws of thermodynamics.

SPEAKER_00 4:41

It's a fight we can't sustain.

SPEAKER_01 4:43

No, we can't just keep laying more copper and installing bigger air conditioners. We are going to hit an energy wall where the global power grid simply cannot generate enough electricity to support the digital economy.

SPEAKER_00 4:55

And society clearly isn't going to stop using AI.

SPEAKER_01 4:58

Right. So how do we escape this physical limitation?

SPEAKER_00 5:01

Aaron Powell Well, we don't stop. We change the medium entirely. And this is where the IOWN Global Forum's radical alternative comes into play.

SPEAKER_01 5:08

Okay, tell us about IOWN.

SPEAKER_00 5:10

So IOWN stands for the Innovative Optical and Wireless Network. Their core architectural concept is the all photonics network, or APN.

SPEAKER_01 5:20

Aaron Powell So this is the shift from electronics to photonics.

SPEAKER_00 5:23

Yes. Instead of forcing electrons through resistant metal wires, the APN uses photons, particles of light traveling over optical fiber.

SPEAKER_01 5:32

Oh, this is where Kawashima's analogy comes in. I loved this part of the text. He says while using electronic wire is like dragging something across a high-resistant surface, sending data over fiber via photonics is like moving over the air.

SPEAKER_00 5:45

It's a great comparison. If we go back to your garden hose analogy, or maybe imagine running through waste deep water photonics, is the equivalent of ice skating on a freshly zambonied rink.

SPEAKER_01 5:54

Oh, I like that. Near frictionless.

SPEAKER_00 5:56

Exactly. The data glides over the glass fiber without the collisions, without the resistance, and critically without generating that massive amount of heat.

SPEAKER_01 6:04

Aaron Powell So the white paper outlines five critical requirements, five pillars that optical networks must evolve to deliver, right? Yes.

SPEAKER_00 6:10

First is even greater bandwidth, but within tight capex targets, meaning without spending billions on new infrastructure.

SPEAKER_01 6:17

Right. What's the second?

SPEAKER_00 6:18

Lower latency across diverse sectors. Things have to be faster.

SPEAKER_01 6:22

Makes sense.

SPEAKER_00 6:23

Third is lower power consumption per bit, which is crucial for sustainability.

SPEAKER_01 6:28

Escaping that energy wall we just talked about.

SPEAKER_00 6:31

Exactly. And the final two are cloud-like agility and greater network robustness and security.

SPEAKER_01 6:37

Okay, wait. I have to stop you there and push back a little bit on behalf of the listener.

SPEAKER_00 6:41

Sure, go ahead.

SPEAKER_01 6:42

So what does this all mean? Because my home internet provider bombards me with flyers every single week, claiming I have a 100% fiber optic network running to my house. We hear about fiber all the time.

SPEAKER_00 6:54

You do?

SPEAKER_01 6:55

How is this APN environment actually different or scaled up from what we already have?

SPEAKER_00 7:00

It is a fundamentally different architecture, and understanding the gap between your home internet and the APN is crucial. Right now, what you have at home is fiber to the premises. Okay. The physical cable on the ground is glass. But the broader network still heavily relies on electronic processing at every major switching point.

SPEAKER_01 7:19

So the light doesn't stay light?

SPEAKER_00 7:20

No, it doesn't. Think of our current internet like a superhighway, where the cars which represent the data can travel at hundreds of miles per hour. But at every single intersection, the drivers are forced to stop, get out, manually translate their map, get into a totally different car, and then accelerate back up to speed.

SPEAKER_01 7:38

Aaron Powell That sounds incredibly inefficient.

SPEAKER_00 7:40

It is. In telecommunications, this is known as OEO conversion, optical to electrical and back to optical. Electronic routers can't read a photon.

SPEAKER_01 7:49

Aaron Powell Ah, so they have to change it back to electricity to know where to send it.

SPEAKER_00 7:53

Aaron Powell Right. At the data center, the light signal has to be converted back into electricity, processed by a silicon chip to read the routing address, and then converted back into a laser beam. Every one of those conversions introduces lag, consumes massive amounts of power, and generates heat.

SPEAKER_01 8:08

Aaron Powell So the all photonics network just removes the intersections.

SPEAKER_00 8:11

Exactly. Kawashima clarifies that this is an open environment where multiple industry players participate directly in the photonic layer. The data travels from end to end as pure light.

SPEAKER_01 8:21

Aaron Powell But how does that actually reach me? My phone doesn't have a glass fiber plugged into it.

SPEAKER_00 8:26

Aaron Powell Right. And that's where Omni's concept of extended network capillarity comes in. The pure optical network branches out incredibly close to the end user to the street corner or the ceiling of an office building.

SPEAKER_01 8:38

Aaron Powell And then what?

SPEAKER_00 8:39

Only at that absolute final millimeter does it use short-range radio access to connect local devices. It handles the massive new capacity over the air for just the last few feet.

SPEAKER_01 8:50

Wow. Okay, here's where it gets really interesting, especially for anyone who cares about urban planning or real estate, because this incredible speed and lower power consumption unlocks something fascinating about where things are physically located.

SPEAKER_00 9:03

It fundamentally changes the map.

SPEAKER_01 9:04

It really does. Kaoshima points out a huge problem in the financial sector right now. Currently, the massive data centers for big banks are clustered right in urban downtown areas.

SPEAKER_00 9:13

Like lower Manhattan or downtown Tokyo.

SPEAKER_01 9:15

Yeah, and that creates a massive density of power demand in a very small, already crowded area.

SPEAKER_00 9:20

Aaron Powell And the reason they are trapped there is purely physics. In the current electronic paradigm, distance equals delay. If a Wall Street trading algorithm is located in Ohio, the milliseconds it takes for those electrons to fight their way back to New York is enough time for a rival bank to beat them to a trade.

SPEAKER_01 9:39

So they literally have to be physically next door to their offices.

SPEAKER_00 9:42

Exactly. But the IOWN solution, the all photonics network, breaks the real estate trap. Because APN offers incredibly high bandwidth and low latency.

SPEAKER_01 9:56

Because the light is so fast, distance no longer creates meaningful latency.

SPEAKER_00 10:00

Right. This is a major geographic shift. Companies can now deploy data centers out into suburban or rural areas.

SPEAKER_01 10:06

This is such an elegant solution. We solve a digital carbon emissions problem by physically moving the buildings out of the city.

SPEAKER_00 10:13

It dramatically lowers the power demand density per area, and it allows them to take advantage of abundant renewable energy out in the countryside. Space is cheap, and the green energy is plentiful.

SPEAKER_01 10:23

Connected by a tether of light back to the city. I love that. But what kind of next generation work will these physically distant, sustainably powered data centers actually be doing, particularly for the financial institutions Kawashima works with?

SPEAKER_00 10:37

Well, this brings us to real world resilience. Let's look at how financial services will utilize this infrastructure. By connecting AI computing resources to service platforms via these low latency connections, banks can instantly analyze vast amounts of transaction logs.

SPEAKER_01 10:52

Oh, like looking for fraud in real time.

SPEAKER_00 10:55

Exactly. This fundamentally changes the game for disaster recovery, operational resiliency, and security. You can block a fraudulent transaction the millisecond a card is swiped, even if the AI server is hundreds of miles away.

SPEAKER_01 11:08

And I want to make sure we connect this back to you, the listener, because the white paper breaks down how this benefits everyone. It's not just banks.

SPEAKER_00 11:14

No, not at all.

SPEAKER_01 11:15

For consumers, it means better performance for recreation, education, and vocation. Imagine seamless virtual reality without any lag. And enterprises get a much better return on investment for their cloud and AI adoption.

SPEAKER_00 11:29

And we shouldn't forget the government angle. For governments, APNs are crucial for maintaining sovereignty, national competitiveness, and running critical infrastructure like defense and healthcare.

SPEAKER_01 11:41

So this all sounds amazing, but I have to ask the million-dollar question.

SPEAKER_00 11:44

Let's hear it.

SPEAKER_01 11:45

When does this sci-fi reality actually happen? Are we talking decades from now?

SPEAKER_00 11:51

Aaron Powell It's actually grounded in a much more realistic timeline. Kawashima notes that feasibility demonstrations have just started.

SPEAKER_01 11:57

Aaron Powell So the physics work and the prototypes are functional right now.

SPEAKER_00 12:00

Aaron Powell Yes. But as he points out, it takes time to prove the commercial operability of this technology for the broader industry. You have to standardize this across telecom giants.

SPEAKER_01 12:10

Aaron Powell So how long until we see it everywhere?

SPEAKER_00 12:12

Aaron Powell He estimates it will take a few years of these focused demonstrations and operational proofs before we see large-scale commercial deployment.

SPEAKER_01 12:20

Aaron Powell Okay, so it's close, but it requires some strategic patience. Aaron Powell Let's briefly recap the journey we've been on today.

SPEAKER_00 12:26

Aaron Powell It's been a lot of ground to cover.

SPEAKER_01 12:28

Aaron Powell It really has. We started with this looming AI power crisis, that staggering growth in power load that threatens our grids. And the solution is replacing those heat-generating resistant electrons with frictionless photons.

SPEAKER_00 12:42

The all photonics network.

SPEAKER_01 12:44

Right. And by doing that, we can literally move our massive data centers out of crowded cities, turning them into rural green energy hubs, all while supercharging financial security.

SPEAKER_00 12:54

It's a remarkable shift. And it leaves us with a final lingering question that builds entirely on this premise.

SPEAKER_01 13:01

Oh, I'm intrigued. What is it?

SPEAKER_00 13:03

Well, think about it. If the Alpha Tonics network means that immense distance no longer creates a lag in data or application performance, how might this change the physical layout of our future cities? If the brain of a business or government could be instantly accessed from hundreds of miles away in a rural area, we might not just decentralize our data centers. We might completely redefine what a headquarters or an urban financial district even needs to be in a digital economy.

SPEAKER_01 13:30

Wow. That is wild to think about. If distance is dead, the whole map changes. Thank you so much for joining us on this deep dive today. Keep asking questions, keep challenging how you think the physical world works, and stay curious. We will catch you next time.