The Physics of Tomorrow’s AI: How LongServing Technology Is Rethinking Computing

For generations, the story of computing has been written in electrons.

Every smartphone, laptop, supercomputer and AI accelerator works by moving electrical signals through microscopic silicon circuits. For decades, that approach has become faster, smaller and more efficient, driving everything from the internet to the rise of artificial intelligence.

Now, however, the rules that shaped modern computing are beginning to change.

Artificial intelligence is placing demands on computing systems unlike anything engineers have faced before. Training large AI models requires enormous computational power, vast amounts of energy and increasingly sophisticated hardware. Every leap in AI capability places another layer of pressure on the infrastructure beneath it.

The challenge is no longer simply to build smarter algorithms.

It is to build computers capable of supporting them.

That question has led researchers around the world to explore an idea that once seemed confined to physics laboratories: what if computers could process information with light instead of electricity?

Among the companies investigating that possibility is LongServing Technology, a Taiwan-based deep technology company founded by Dr. Ko-Cheng Fang. Its work sits at the intersection of optics, materials science and computer engineering, where one of the world’s oldest scientific phenomena—light—is being reconsidered as the foundation of tomorrow’s computing systems.

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.

The Problem Isn’t AI. It’s What AI Demands.

Most discussions about artificial intelligence focus on software.

New language models. Better image generators. Smarter assistants.

Yet every improvement in AI depends on hardware performing trillions of calculations with extraordinary speed and precision.

Today’s processors achieve this using billions of transistors fabricated on silicon chips. That architecture has served computing remarkably well, but it comes with growing constraints.

Electronic circuits consume significant amounts of electricity. They generate heat that must be removed through increasingly complex cooling systems. Manufacturing becomes more difficult as semiconductor features shrink towards atomic dimensions.

None of these problems prevent AI from advancing.

But together they make every new generation of computing more challenging—and more expensive—to build.

For researchers, the question is becoming increasingly important:

Can computing continue evolving by refining silicon alone, or will entirely new architectures eventually be required?

LongServing’s Scientific Bet

LongServing Technology believes the answer may lie in photonic computing.

Instead of using electrons to move information through circuits, photonic systems use photons.

The physics behind the idea is well understood.

Photons travel rapidly, generate far less heat than moving electrical current and offer the possibility of transporting information with exceptional efficiency. For future AI systems that demand enormous computational throughput, those characteristics are particularly attractive.

The science, however, has never been the greatest obstacle.

Engineering has.

Unlike electrical current, which can be routed through highly complex pathways, light naturally prefers travelling in straight lines.

That seemingly simple property becomes a major challenge when building processors, where information must constantly change direction within microscopic circuits.

Controlling light at that scale has remained one of photonic computing’s defining engineering problems.

Why X-Photon Matters

This is where LongServing’s latest research enters the picture.

The company has developed X-Photon, a proprietary optical material designed to guide photons through nanoscale pathways while allowing controlled 90-degree directional changes inside the material itself.

To someone outside engineering, changing the direction of light may appear to be a small achievement.

In reality, it represents one of the practical challenges that has limited integrated photonics for years.

Every future optical processor will require light to move through increasingly complex circuit layouts while maintaining speed and signal quality. If photons cannot be routed reliably, scaling photonic processors becomes extraordinarily difficult.

According to LongServing, X-Photon has been engineered specifically to address this challenge.

If the approach proves scalable under commercial manufacturing conditions, it could become one of the enabling technologies needed for more sophisticated photonic processors.

Building a Platform, Not a Prototype

Perhaps what distinguishes LongServing Technology is that it is not treating X-Photon as an isolated invention.

The company is attempting to build an entire photonic computing platform.

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

That systems-level thinking reflects an important lesson from computing history.

Silicon did not transform the world because of one breakthrough.

It succeeded because advances in materials, chip design, manufacturing, software, packaging and global supply chains evolved together.

LongServing appears to be pursuing a similar philosophy.

Rather than developing a single component, it is working towards the broader ecosystem that photonic computing would ultimately require.

Science Meets Commercial Reality

Scientific breakthroughs rarely become technologies overnight.

Between discovery and widespread adoption lies a long process of engineering refinement, manufacturing validation and commercial execution.

Recognising this, LongServing recently announced a US$500 million financing initiative, based on a stated valuation of US$2.5 billion. According to the company, the investment is intended to expand photonic manufacturing capability, strengthen research programmes and accelerate commercial development.

For students of science, this illustrates an important reality.

Modern innovation depends as much on translating research into industrial capability as it does on making the original discovery.

The laboratory is only the beginning.

The Next Scientific Frontier

Whether photonic computing eventually replaces conventional semiconductor technology remains uncertain.

Silicon continues to evolve, and decades of manufacturing expertise make it one of the most successful engineering platforms ever developed.

Yet history also shows that transformative advances often begin when researchers question assumptions that once seemed permanent.

LongServing Technology is part of that tradition.

Its work does not suggest that silicon’s era is ending tomorrow. Instead, it reflects a growing scientific effort to explore what comes after the limits of conventional electronics.

For students studying physics, materials science, electronics, nanotechnology or computer engineering, that may be the most important lesson of all.

The future of artificial intelligence will not be determined only by better software.

It will also depend on the scientists and engineers willing to rethink the physical foundations of computation itself.

LongServing Technology’s research is one example of that search—a reminder that some of tomorrow’s biggest breakthroughs may begin not with a new algorithm, but with a new understanding of how light can be harnessed to compute.

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