Scott Galloway:
“Nuclear power may be the worst-managed brand in history”
By exploring and sharing insights from this field,
you’re not just reporting
on a scientific breakthrough—you’re telling the story of how humanity could solve
one of its greatest challenges and redefine its future.
you’re not just reporting
on a scientific breakthrough—you’re telling the story of how humanity could solve
one of its greatest challenges and redefine its future.
For decades, nuclear power has played a vital role in global energy production, yet concerns over waste management and safety persist. Now, fusion energy is emerging as a game-changer. Unlike fission, which splits atoms apart, fusion unites atomic nuclei in an ultra-hot, electrically charged state of matter called plasma—the same process that fuels the sun and stars.
But harnessing fusion energy comes with significant challenges. The biggest challenge is confining and sustaining plasma. At extreme temperatures exceeding those in the sun’s core, plasma becomes highly dynamic and difficult to maintain. Developing reliable methods to keep it confined and stable is essential for achieving practical fusion energy.
When scientists succeed, fusion will provide a scalable, clean energy source capable of meeting global demand for generations. As journalists, your role is crucial in shaping public understanding of fusion’s promise.
But harnessing fusion energy comes with significant challenges. The biggest challenge is confining and sustaining plasma. At extreme temperatures exceeding those in the sun’s core, plasma becomes highly dynamic and difficult to maintain. Developing reliable methods to keep it confined and stable is essential for achieving practical fusion energy.
When scientists succeed, fusion will provide a scalable, clean energy source capable of meeting global demand for generations. As journalists, your role is crucial in shaping public understanding of fusion’s promise.
Founder & CEO, Next Step Fusion
Aleksei Zolotarev
“At Next Step Fusion, we believe that recent breakthroughs and upcoming advancements will propel the fusion energy industry forward, delivering safe and affordable power for humanity. This transformation is already underway and will accelerate in the coming years.
With collaboration between scientists and commercial companies, fusion could transition from a breakthrough to a practical energy source within the next decade, making clean, stable, and abundant energy as common as the internet—something that, just a few decades ago, seemed like science fiction.”
With collaboration between scientists and commercial companies, fusion could transition from a breakthrough to a practical energy source within the next decade, making clean, stable, and abundant energy as common as the internet—something that, just a few decades ago, seemed like science fiction.”
Mastering Plasma Control:
The Key to Commercial Fusion
The Key to Commercial Fusion
Tokamaks are the most researched fusion approach, using strong magnetic fields to confine hot plasma in a toroidal chamber, enabling fusion conditions.
Since 1958, over 200 tokamaks have been built, with around 60 currently operating. Book design is the art of incorporating the content, style, format, design, and sequence of the various components of a book into a coherent whole. In the words of Jan Tschichold, book designer, "methods and rules upon which it is impossible to improve, have been developed over centuries. To produce perfect books, these rules have to be brought back to life and applied."
The majority of fusion researchers focus on tokamaks, reflecting confidence in this method. Challenges in plasma stability, turbulence, and steady-state operation remain, but advanced simulations, digital twins, and AI-driven controls are paving the way toward viable fusion power.
Controlling plasma in tokamaks is very difficult because of its chaotic nature and the extreme speed of events and processes. Despite significant progress, tokamak control remains one of the biggest challenges. That’s why Next Step Fusion is focused on developing technologies and products for reliable plasma control that suit not only current and next-generation research tokamaks but also future fusion power plants.
Controlling plasma in tokamaks is very difficult because of its chaotic nature and the extreme speed of events and processes. Despite significant progress, tokamak control remains one of the biggest challenges. That’s why Next Step Fusion is focused on developing technologies and products for reliable plasma control that suit not only current and next-generation research tokamaks but also future fusion power plants.
Our Partners
The major blocks of our ecosystem are:
NSFsim Advanced Tokamak Simulator
NSFsim is a powerful simulation tool that helps researchers and engineers predict and control the behavior of hot plasma inside tokamaks. It uses precise geometry of each device to model how plasma interacts with magnetic fields, electrical circuits, and surrounding structures.
At Next Step Fusion, we use NSFsim alongside other advanced tools to conduct feasibility studies and develop preliminary designs for new tokamaks, helping to shape the future of fusion energy.
At Next Step Fusion, we use NSFsim alongside other advanced tools to conduct feasibility studies and develop preliminary designs for new tokamaks, helping to shape the future of fusion energy.
A free web-based tool designed to democratize access to advanced tokamak simulations, enable collaborative research in fusion science, and enhance plasma physics and fusion engineering education.
FTP allows researchers, educators, and students to use pre-built digital replicas of tokamaks, enabling precise simulations, exploration of machine learning models, visualization of plasma dynamics, and flexible data management.
FTP allows researchers, educators, and students to use pre-built digital replicas of tokamaks, enabling precise simulations, exploration of machine learning models, visualization of plasma dynamics, and flexible data management.
Plasma Control System
A modern industrial PCS for tokamaks, stellarators, and future power plants, built on the plasma state — key parameters defining plasma behavior for real-time control.
The PCS has simple yet flexible architecture with clear separation of control layers, and combines conventional and ML-based control methods to ensure reliability and efficiency.
The PCS has simple yet flexible architecture with clear separation of control layers, and combines conventional and ML-based control methods to ensure reliability and efficiency.
If you need a comment, just fill out the form—
we’ll get back to you within 24 hours!
we’ll get back to you within 24 hours!
We’re always open and happy to talk about plasma control and tokamak management systems. Subscribe to our educational newsletter, delivered once a month—no spam, just expert insights. Each issue covers key industry events, notable scientists, and emerging trends, keeping you informed and ahead of the curve in industry news.
Technical Questions & Answers
Fusion energy is advancing rapidly, and with it come complex technical challenges. We know journalists need clear, reliable insights to navigate this evolving field. Below, we answer key questions about our technology, approach, and the future of fusion, providing you with the context needed to report on the next big breakthroughs in energy.
Where does your expertise in fusion come from?
Does Next Step Fusion only specialize in plasma physics and fusion research?
Why are you developing your own Plasma Control System?
What else do you do in fusion besides developing a Plasma Control System?
Why is NSFsim so important?
How do you achieve accurate simulation of different tokamaks?
What is the Tokamak Concept Feasibility Study?
How do you design tokamaks for your customers?
What AI/ML solutions do you develop?
Why did you create the Fusion Twin Platform?
Which tokamaks already have digital replicas on the platform?
Where does your expertise in fusion come from?
Our team has extensive experience working on major fusion projects and devices like ITER, KTM, T-10, T-15MD, COMPASS, DIII-D, and ISTTOK.
We are also supported by a network of consultants, partners, and collaborators, with whom we design, test, and implement innovative solutions for the fusion industry. For example we collaborate with a range of research organizations, such as Columbia University (USA), DIII-D National Fusion Facility (USA), University of California San Diego (USA), Instituto Superior Técnico (Portugal), University of Sevilla (Spain), Tuscia University (Italy), Consorzio RFX (Italy), and a few others.
In order to stay closely connected to the latest scientific advancements and technological developments in the fusion field we participate in plasma physics and fusion conferences, run collaborative projects and experiments, sponsor PhD positions, and much more.
We are also supported by a network of consultants, partners, and collaborators, with whom we design, test, and implement innovative solutions for the fusion industry. For example we collaborate with a range of research organizations, such as Columbia University (USA), DIII-D National Fusion Facility (USA), University of California San Diego (USA), Instituto Superior Técnico (Portugal), University of Sevilla (Spain), Tuscia University (Italy), Consorzio RFX (Italy), and a few others.
In order to stay closely connected to the latest scientific advancements and technological developments in the fusion field we participate in plasma physics and fusion conferences, run collaborative projects and experiments, sponsor PhD positions, and much more.
Does Next Step Fusion only specialize in plasma physics and fusion research?
No, our expertise goes beyond plasma physics and fusion research. At Next Step Fusion, we combine deep knowledge of fusion science with strong capabilities in data science, machine learning, software development, analytics, and business strategy. Our team includes experts in AI-driven plasma control, large-scale simulations, and digital twin technologies, enabling us to develop cutting-edge solutions for optimizing tokamak operations and advancing fusion research. Additionally, our experience in software engineering and data infrastructure allows us to build scalable platforms like the Fusion Twin Platform, while our business and strategic expertise ensures we position these innovations effectively within the global fusion ecosystem.
Why are you developing your own Plasma Control System?
Because existing scientific systems are not suitable for commercial companies to use in controlling energy-producing reactors. Current control systems and methods are designed for short pulses in tokamaks that are small and weak compared to future reactors. That’s why we are developing our own system, combining the best practices from conventional control with modern AI/ML technologies.
What else do you do in fusion besides developing a Plasma Control System?
Besides developing a Plasma Control System, we offer the fusion industry:
- Advanced tokamak simulation services using NSFsim, our advanced tokamak simulator
- Tokamak concept feasibility studies
- Tokamak optimization and design projects
- AI/ML solutions
Why is NSFsim so important?
NSFsim is crucial because it provides a precise environment for simulating tokamaks. It integrates 2D Grad-Shafranov solvers, 1D transport solvers, and various other algorithms for self-consistent calculations. Additionally, it has undergone rigorous validation against real experimental data from more than 20 tokamaks, ensuring its accuracy and reliability.
Researchers and engineers use NSFsim to conduct advanced simulations, train and test reinforcement learning-based control agents, and support the development of conventional controllers, such as PID. By modeling how plasma interacts with magnetic fields, electrical circuits, and surrounding structures using precise device geometry, NSFsim contributes to improving reactor performance and stability.
Researchers and engineers use NSFsim to conduct advanced simulations, train and test reinforcement learning-based control agents, and support the development of conventional controllers, such as PID. By modeling how plasma interacts with magnetic fields, electrical circuits, and surrounding structures using precise device geometry, NSFsim contributes to improving reactor performance and stability.
As a key research tool, NSFsim supports the development of technologies essential for advancing commercial fusion energy. Its ability to simulate different operational scenarios helps optimize plasma control strategies, diagnose potential instabilities, and enhance overall tokamak performance.
At Next Step Fusion, NSFsim is at the core of everything we do. We leverage it alongside other advanced tools to conduct feasibility studies and develop preliminary designs for new tokamaks, shaping the future of fusion energy. We also collaborate with researchers and industry partners to integrate AI-driven control solutions and explore novel reactor configurations, driving innovation in fusion technology.
At Next Step Fusion, NSFsim is at the core of everything we do. We leverage it alongside other advanced tools to conduct feasibility studies and develop preliminary designs for new tokamaks, shaping the future of fusion energy. We also collaborate with researchers and industry partners to integrate AI-driven control solutions and explore novel reactor configurations, driving innovation in fusion technology.
If you need a comment, just fill out the form—
we’ll get back to you within 24 hours!
we’ll get back to you within 24 hours!
We’re always open and happy to talk about plasma control and tokamak management systems. Subscribe to our educational newsletter, delivered once a month—no spam, just expert insights. Each issue covers key industry events, notable scientists, and emerging trends, keeping you informed and ahead of the curve in industry news.
How do you achieve accurate simulation of different tokamaks?
Creating accurate simulations of tokamak behavior requires a precise mathematical representation of real devices. This is where tokamak digital replicas come in—unique virtual models of each machine. While we refer to them as digital replicas, they are sometimes also called digital twins or device configurations.
These digital replicas incorporate key geometrical and electrical properties, including magnetic coils, the vacuum vessel, and passive conducting structures.
Once a tokamak digital replica is complete, we validate its accuracy against experimental data from our partner scientists. This ensures high-fidelity simulations that not only replicate real plasma discharges but also provide a controlled environment for training reinforcement learning agents to improve plasma control.
These digital replicas incorporate key geometrical and electrical properties, including magnetic coils, the vacuum vessel, and passive conducting structures.
Once a tokamak digital replica is complete, we validate its accuracy against experimental data from our partner scientists. This ensures high-fidelity simulations that not only replicate real plasma discharges but also provide a controlled environment for training reinforcement learning agents to improve plasma control.
What is the Tokamak Concept Feasibility Study?
The Tokamak Concept Feasibility Study is a comprehensive evaluation of the scientific, engineering, and economic viability of building a tokamak-based fusion system. At Next Step Fusion, we use NSFsim and our other advanced tools to conduct such studies for our customers—companies aiming to develop new conventional or spherical tokamaks for various applications.
How do you design tokamaks for your customers?
We design tokamaks for our customers because we believe in the tokamak concept as the most researched and developed approach to fusion energy. Our goal is to increase the number of private tokamaks worldwide, as we believe private companies will achieve fusion energy production faster.
Beyond energy production, we see tokamaks playing a crucial role in other applications, such as neutron sources for medical isotope production, propulsion systems, industrial heat generation, and more.
That is why we support early-stage tokamak companies by enabling them to quickly start with a Tokamak Concept Feasibility Study and, once completed, move forward with the preliminary design of their tokamak. This includes defining key design parameters, assessing mechanical, electromagnetic, and neutron loads, determining tokamak geometry and subsystems, conducting disruption simulations, and much more.
Beyond energy production, we see tokamaks playing a crucial role in other applications, such as neutron sources for medical isotope production, propulsion systems, industrial heat generation, and more.
That is why we support early-stage tokamak companies by enabling them to quickly start with a Tokamak Concept Feasibility Study and, once completed, move forward with the preliminary design of their tokamak. This includes defining key design parameters, assessing mechanical, electromagnetic, and neutron loads, determining tokamak geometry and subsystems, conducting disruption simulations, and much more.
What AI/ML solutions do you develop?
We develop AI/ML-powered solutions for fusion, recognizing their immense potential in handling the chaotic nature of plasma, the complexity of underlying physical processes, and the intricate designs of tokamak devices. At Next Step Fusion, we believe in combining conventional methods with AI/ML to deliver fast and reliable simulations, control systems, design tools, and other critical services.
One of our biggest focuses is applying reinforcement learning (RL) to plasma control. This approach, which was first demonstrated by DeepMind and the TCV tokamak, involves training an RL agent to optimize the magnetic control of plasma in real time. Unlike traditional PID-based controllers, RL agents can learn complex, nonlinear control strategies directly from data, enabling them to stabilize plasma in a more adaptive and robust way.
One of our biggest focuses is applying reinforcement learning (RL) to plasma control. This approach, which was first demonstrated by DeepMind and the TCV tokamak, involves training an RL agent to optimize the magnetic control of plasma in real time. Unlike traditional PID-based controllers, RL agents can learn complex, nonlinear control strategies directly from data, enabling them to stabilize plasma in a more adaptive and robust way.
We successfully reproduced this approach on the DIII-D tokamak, where our RL-based control model demonstrated the ability to maintain plasma stability in L-mode, H-mode, and H-L transition. We are now expanding our work by testing it on other tokamaks, increasing the number of plasma parameters that RL agents can control, and ultimately providing RL-driven plasma control as a service for tokamak facilities.
Another key direction is the development of fast surrogate models. These ML-driven models are trained on historical datasets or a combination of historical and synthetic data to replace slow, first-principles scientific codes. Some traditional simulations can take weeks to compute a single time step on high-performance computing (HPC) clusters. By leveraging surrogate models, we significantly accelerate these computations while maintaining high accuracy, making real-time analysis and rapid design iteration possible for tokamak research and development.
Another key direction is the development of fast surrogate models. These ML-driven models are trained on historical datasets or a combination of historical and synthetic data to replace slow, first-principles scientific codes. Some traditional simulations can take weeks to compute a single time step on high-performance computing (HPC) clusters. By leveraging surrogate models, we significantly accelerate these computations while maintaining high accuracy, making real-time analysis and rapid design iteration possible for tokamak research and development.
If you need a comment, just fill out the form—
we’ll get back to you within 24 hours!
we’ll get back to you within 24 hours!
We’re always open and happy to talk about plasma control and tokamak management systems. Subscribe to our educational newsletter, delivered once a month—no spam, just expert insights. Each issue covers key industry events, notable scientists, and emerging trends, keeping you informed and ahead of the curve in industry news.
Why did you create the Fusion Twin Platform?
Running complex tokamak simulations was once limited to well-funded university teams with high-performance computing. The Fusion Twin Platform changes that, enabling small research groups and independent enthusiasts to run advanced fusion simulations, analyze experimental data, and contribute to discoveries—all from a web-based interface.
Users can configure simulations, visualize results with predefined graphs, and collaborate seamlessly. Unlike static images or offline files, the platform offers real-time, dynamic simulations, making fusion research more accessible.
Users can configure simulations, visualize results with predefined graphs, and collaborate seamlessly. Unlike static images or offline files, the platform offers real-time, dynamic simulations, making fusion research more accessible.
Beyond simulations, the platform advances fusion education by providing students and researchers hands-on experience with real data and plasma modeling, bridging theory and practice.
Key Features:
Key Features:
- Digital replicas of real tokamaks (DIII-D, ISTTOK, SMART, NSF NTT, etc.)
- Highly customizable plasma equilibrium modeling
- Fast visualization of simulation results via hundreds of pre-configured graphs and a customizable graph editor
- Data management tools for uploading experimental data, downloading simulated data, exporting customizable graph images, workspace sharing, and public link generation
- Fully cloud-based, no local installation needed
Which tokamaks already have digital replicas on the platform?
The platform already includes digital replicas of the DIII-D, ISTTOK, SMART, and NSF NTT tokamaks. Other tokamaks are coming soon.
DIII-D is one of the world's leading tokamaks, located in San Diego, California, United States, and operated by General Atomics with support from the U.S. Department of Energy (DOE). It plays a crucial role in research supporting the development of future fusion reactors, including ITER. Next Step Fusion is a registered user of the DIII-D Program and an active collaborator, with several completed projects and published or in-review papers.
ISTTOK (Instituto Superior Técnico TOKamak) is a compact tokamak located in Lisbon, Portugal. It is operated by the Instituto de Plasmas e Fusão Nuclear (IPFN) at Instituto Superior Técnico (IST), which is part of the University of Lisbon.
ISTTOK (Instituto Superior Técnico TOKamak) is a compact tokamak located in Lisbon, Portugal. It is operated by the Instituto de Plasmas e Fusão Nuclear (IPFN) at Instituto Superior Técnico (IST), which is part of the University of Lisbon.
ISTTOK is used for research in plasma physics, diagnostics, and the development of new plasma control techniques. Next Step Fusion is a partner of IPFN, working to test our reinforcement learning (RL) controllers on ISTTOK.
SMART (Small Aspect Ratio Tokamak) – is an experimental spherical tokamak designed, built, and operated by the Plasma Science and Nuclear Fusion Technology Laboratory at the University of Seville in Spain. It is unique in its ability to shape plasma in various forms, including negative triangularity. We are extremely interested in negative triangularity and in helping SMART become a leader in NT research by providing the best plasma control solutions.
NSF NTT exists only in digital form and represents the design of a negative triangularity tokamak developed jointly by Next Step Fusion and Columbia University. This work was dedicated to establishing best practices for tokamak feasibility studies, preliminary design, simulation and design tool adoption, and building relationships within the supply chain.
SMART (Small Aspect Ratio Tokamak) – is an experimental spherical tokamak designed, built, and operated by the Plasma Science and Nuclear Fusion Technology Laboratory at the University of Seville in Spain. It is unique in its ability to shape plasma in various forms, including negative triangularity. We are extremely interested in negative triangularity and in helping SMART become a leader in NT research by providing the best plasma control solutions.
NSF NTT exists only in digital form and represents the design of a negative triangularity tokamak developed jointly by Next Step Fusion and Columbia University. This work was dedicated to establishing best practices for tokamak feasibility studies, preliminary design, simulation and design tool adoption, and building relationships within the supply chain.
OUR NEWS
Let’s connect and stay in touch!
Let’s talk about fusion.
If you’re covering the future of energy, plasma physics, or AI-driven control systems, we’re happy to share insights, data, and expertise. Feel free to reach out—we’re always open to meaningful discussions.
If you’re covering the future of energy, plasma physics, or AI-driven control systems, we’re happy to share insights, data, and expertise. Feel free to reach out—we’re always open to meaningful discussions.