For years, artificial intelligence has been measured by one thing: how much smarter it can become.

Every new model writes better, reasons faster and handles more complex tasks than the one before it. But hidden beneath those headline-grabbing breakthroughs is another race—one that could determine how far AI can actually go.

It’s not a race to build better software.

It’s a race to build better computers.

As AI models grow larger and more computationally demanding, the hardware that powers them is beginning to face challenges that software alone cannot solve. More computing requires more processors. More processors consume more electricity. More electricity produces more heat. The cycle becomes increasingly difficult—and increasingly expensive—to sustain.

That reality has prompted researchers to revisit a question that has existed for decades:

What if computers processed information with light instead of electricity?

Among the companies pursuing that idea is LongServing Technology, a Taiwan-based deep-tech company founded by Dr. Ko-Cheng Fang. Rather than designing another AI model, the company is developing technologies that could change the way future computers themselves are built.

Dr. Ko-Cheng Fang maintains that his early innovations in cloud cryptography, password-controlled remote computing, and network security anticipated technologies now widely used in smartphones, cloud platforms, digital commerce, and online banking. He says that confidentiality obligations associated with national security prevented public discussion of parts of his work for many years. Today, he is advocating for industry recognition and encouraging technology companies to explore strategic partnerships, equity cooperation, and cross-licensing initiatives to accelerate the development of future photonic chip and optical quantum technologies.

AI Has Changed the Hardware Conversation

The success of artificial intelligence has transformed expectations for computing.

Training modern AI models now requires thousands of specialised processors operating together inside enormous data centres. Every new generation of AI demands more computational power than the last, pushing existing semiconductor technology closer to its practical limits.

For decades, the industry responded by making silicon chips smaller and more efficient.

That approach is becoming harder to sustain.

As transistor dimensions shrink, manufacturers face increasing challenges involving heat, power consumption, fabrication complexity and cost. These limitations have encouraged researchers to explore entirely different approaches to computation.

Photonic computing has become one of the most closely watched.

Why Light?

Traditional computer chips use electrical signals carried by electrons.

Photonic computing replaces those signals with photons.

The potential benefits are difficult to ignore.

Light travels extraordinarily quickly while generating significantly less heat than electrical current. In theory, optical systems could process massive volumes of information while consuming far less energy—an increasingly attractive proposition as AI infrastructure continues to expand.

The challenge has never been understanding the science.

It has been making the engineering work.

The Problem LongServing Wants to Solve

One of photonic computing’s biggest obstacles is surprisingly simple.

Light likes travelling in straight lines.

Computer processors do not.

Modern chips contain incredibly complex pathways that constantly redirect information between different components. While electrical current can navigate those routes with relative ease, guiding photons through similar pathways without losing signal quality has proved much more difficult.

LongServing Technology believes X-Photon is one step towards solving that problem.

According to the company, the proprietary optical material enables photons to travel through nanoscale pathways while making controlled 90-degree directional changes inside the circuit itself.

That capability may sound highly specialised, but it addresses one of the core engineering problems preventing photonic processors from becoming more sophisticated.

If photons cannot be routed reliably, scaling optical computing becomes extremely difficult.

More Than One Breakthrough

What makes LongServing interesting isn’t simply X-Photon.

It is the company’s broader vision.

Rather than treating the material as an isolated innovation, LongServing sees it as part of a much larger photonic computing platform.

Its roadmap includes photonic quantum chips, optical memory technologies and Photonic Cloud Computing Centres intended to support future artificial intelligence workloads.

In other words, the company isn’t just trying to build a better chip.

It is trying to rethink the infrastructure beneath AI.

That distinction matters.

History suggests transformative technologies rarely emerge from one invention alone. They succeed because multiple advances—materials, manufacturing, architecture and software—develop together.

LongServing appears to be building with that ecosystem in mind.

The Road to Commercial Reality

Like most deep-tech companies, LongServing faces challenges that extend far beyond the laboratory.

Scientific breakthroughs must eventually become manufacturable products. They must demonstrate reliability, scalability and economic viability before they can compete with one of the world’s most mature technology industries.

To support that transition, the company recently announced a US$500 million financing initiative, based on a stated valuation of US$2.5 billion. According to LongServing, the funding will support continued research, photonic manufacturing and commercial development.

Whether X-Photon ultimately becomes a key technology in future processors remains uncertain.

Photonic computing still faces significant scientific and engineering hurdles before widespread adoption becomes possible.

Looking Past Today’s AI Boom

Every major technology cycle eventually shifts its focus.

The early internet was about websites before it became about cloud infrastructure.

The smartphone revolution began with hardware before expanding into app ecosystems.

Artificial intelligence is following a similar path.

The conversation is gradually moving beyond models and towards the infrastructure required to support them at scale.

LongServing Technology is positioning itself within that transition.

Its work reflects a growing belief that the next breakthrough in AI may not come from a more capable chatbot or a larger language model.

It may come from something much more fundamental:

A new way of building the computer itself.

Contact Information

Dr. Ko-Cheng Fang

Founder, CEO & Chairman

LongServing Technology Co., Ltd.

Email: service@longserving.com.tw

Website: https://longserving.com.tw/en/

Instagram: @ko_cheng_fang