Global race to develop Exascale Supercomputers, the key to scientific revolutions to military and strategic superiority

Rajesh Uppal 3 weeks ago AI & IT, Geopolitics, Strategy & Technological Rivalries, Industry & Market Dynamics Comments Off on Global race to develop Exascale Supercomputers, the key to scientific revolutions to military and strategic superiority 1,084 Views

Exascale supercomputers are transforming science, defense, and global power—defining who will lead the 21st century in code, silicon, and strategy.

In the 21st century, computing power has become a cornerstone of national competitiveness. From unraveling the complexities of the universe to designing next-generation weapons, the ability to process vast quantities of data in real time is no longer a luxury—it’s a necessity. At the forefront of this technological revolution is Exascale computing, a domain in which countries are now locked in a high-stakes race for dominance.

What Is Exascale Computing, and Why Does It Matter?

Exascale computing represents a monumental leap in computational power, operating roughly 1,000 times faster than petascale supercomputers. Defined as sustaining one exaflop of 64-bit performance on real-world applications, these machines can perform a quintillion calculations per second. To visualize this scale, if every human on Earth calculated at one operation per second, it would take four years to match the work of a single exascale system in just one second. This extraordinary capability transforms what scientists and engineers can simulate, model, and predict, opening doors to research that was previously inconceivable.

The implications extend far beyond raw speed. Exascale systems are driving breakthroughs in genomics, drug discovery, materials science, climate modeling, and artificial intelligence, while also enabling virtual testing in nuclear physics and national defense. For nations, exascale mastery is not just a matter of technological prestige—it is a strategic imperative. Controlling these computational engines defines the frontier of scientific innovation, economic competitiveness, and military superiority, setting the stage for dominance in the 21st century.

Why the Global Race?

The global race to develop and deploy exascale supercomputers is far more than a technological milestone—it is a geopolitical and economic contest with profound implications.

Scientific Power: Dominance in exascale computing gives nations the ability to simulate complex systems with unprecedented fidelity. From modeling protein folding for drug discovery to exploring quantum interactions or simulating the dynamics of black holes, exascale systems enable insights that were previously impossible. These capabilities accelerate scientific breakthroughs, pushing the boundaries of knowledge in fields ranging from climate science to materials engineering.

National Security Advantage: In defense and strategic planning, exascale computers are invaluable. They facilitate virtual nuclear testing, advanced cryptographic development, and the design and simulation of next-generation missile defense systems—all without the political and environmental risks of real-world experimentation. By simulating complex scenarios at scale, nations can anticipate threats and optimize defense strategies with unmatched precision.

Economic and Industrial Transformation: Exascale computing unlocks transformative potential for industries across the board. Pharmaceutical companies can accelerate drug discovery, energy firms can optimize fusion and renewable energy systems, and manufacturers can refine materials and supply chains with AI-driven modeling. Logistics, finance, and AI-driven innovation all benefit from the ability to process trillions of calculations per second, creating a competitive advantage in the global economy.

Technological Sovereignty: Beyond science and economics, the race is about strategic autonomy. Nations that develop domestic exascale infrastructure can reduce dependence on foreign HPC systems, ensuring control over critical digital assets and maintaining resilience against supply chain disruptions or geopolitical tensions. Achieving technological sovereignty ensures that a nation’s most sensitive scientific and defense workloads remain under its own control.

In essence, the exascale race is a convergence of science, strategy, and sovereignty. Success in this arena defines not only computational supremacy but also a nation’s position in the 21st-century landscape of innovation, security, and economic influence.

Leading Players in the Exascale Race

The exascale computing race has evolved into a multifaceted competition where nations vie not only for computational supremacy but also for technological sovereignty, energy efficiency, and strategic advantage. The United States, China, Japan, and the European Union are at the forefront, each employing unique strategies that blend scientific ambition with geopolitical objectives.

United States: A Layered Exascale Ecosystem

The United States continues to lead in exascale computing with a robust and diversified portfolio of systems. At the pinnacle is El Capitan, located at Lawrence Livermore National Laboratory (LLNL). As of early 2025, El Capitan achieved a peak performance of 2.79 exaflops, making it the world’s most powerful supercomputer. It utilizes a heterogeneous architecture combining AMD’s Instinct MI300A Accelerated Processing Units (APUs) with Cray’s Slingshot interconnect, optimized for both simulation and AI workloads. This system supports critical national security applications, including nuclear stockpile stewardship and advanced materials research.

Following El Capitan, Aurora at Argonne National Laboratory and Frontier at Oak Ridge National Laboratory further solidify the U.S. leadership. Aurora, operational since 2025, integrates Intel’s Sapphire Rapids CPUs with Ponte Vecchio GPUs, facilitating AI-driven scientific research across various disciplines. Frontier, which became the first exascale supercomputer in 2022, combines AMD’s EPYC CPUs with Instinct MI250X GPUs, achieving approximately 21 megawatts of energy usage with high efficiency for large-scale AI-driven scientific workloads.

China: Strategic Self-Reliance and Dual-Use Applications

China’s exascale endeavors are marked by a focus on technological independence and dual-use applications that serve both civilian and military purposes. The Sunway Oceanlite system, believed to achieve 1.9 exaflops in internal benchmarks, employs the domestically developed SW26010-Pro processors and the Sunway RaiseOS operating system. This system is utilized for climate modeling, electromagnetic simulations, and stealth aircraft design. Despite not participating in international benchmarks like the TOP500, China’s commitment to exascale computing is evident through its substantial investments and advancements in domestic semiconductor technologies.

The Tianhe-3 system, another cornerstone of China’s exascale strategy, is reported to achieve 1.7 exaflops peak and 1.3 exaflops sustained performance. It utilizes Phytium ARM-based processors and custom accelerators, supporting applications in aerodynamic simulations and nuclear reactor design. These systems underscore China’s emphasis on self-reliance and the integration of supercomputing capabilities into its broader technological and defense strategies.

Japan: Advancing Towards Zettascale Computing

Japan’s exascale ambitions are embodied in the upcoming FugakuNEXT, a collaboration between RIKEN, Fujitsu, and Nvidia. Scheduled for deployment around 2030, FugakuNEXT aims to achieve a peak performance of 600 exaflops in FP8 sparse precision, positioning it as the world’s first zettascale supercomputer. The system will integrate Fujitsu’s MONAKA-X ARM CPUs with Nvidia GPUs and NVLink Fusion interconnects, emphasizing energy efficiency with a projected power consumption of 40 megawatts. FugakuNEXT is designed to support AI-HPC hybrid workloads, including climate science, drug discovery, and disaster resilience, while also strengthening Japan’s semiconductor sector and global technological leadership.

European Union: JUPITER and the Path to Digital Sovereignty

The European Union has marked a significant milestone with the inauguration of JUPITER, its first exascale supercomputer, at the Forschungszentrum Jülich in Germany. JUPITER achieves a peak performance of 1 exaflop and is powered by Nvidia’s Grace Hopper Superchips and Eviden’s BullSequana XH3000 liquid-cooled architecture. Notably, JUPITER is the most energy-efficient among the world’s top supercomputers, operating entirely on renewable energy and repurposing waste heat for campus heating needs. This system is poised to advance research in climate modeling, AI applications in European languages, neurodegenerative diseases, and retroviruses, while reinforcing Europe’s digital sovereignty and scientific innovation.

Global Trends, Emerging Challenges, and Geopolitical Implications

Across the leading nations, exascale systems share a set of advanced technical characteristics. Modern exascale machines rely on heterogeneous architectures that combine high-performance CPUs with GPUs or APUs, supported by high-bandwidth memory (HBM3 or higher) and ultra-fast interconnects to minimize latency across millions of cores. Energy efficiency is a critical concern, prompting innovations such as liquid immersion cooling, direct-to-chip cooling, and adaptive power management. AI workloads are increasingly integrated alongside traditional simulations, allowing supercomputers to tackle trillion-parameter deep learning models while simultaneously running high-fidelity physics simulations or climate modeling.

The next frontier is zettascale computing, projected to emerge around 2030. Achieving systems capable of 1,000 exaflops introduces formidable challenges in semiconductor fabrication, software scalability, and sustainable energy management. Scaling applications to efficiently utilize millions of cores requires re-architecting legacy scientific codes, developing new parallel programming models, and ensuring fault-tolerant execution over millions of compute nodes. At the same time, energy consumption and thermal management remain critical, with zettascale machines expected to demand tens of megawatts even with aggressive efficiency improvements.

Geopolitical considerations now strongly influence exascale development. The U.S. CHIPS and Science Act and similar export controls have intensified global competition, especially between the U.S. and China, encouraging countries to prioritize technological self-reliance. China has poured over $220 billion into domestic chip fabrication and HPC research under its “Big Fund” initiative, pushing forward processors like the Shenwei SW40000 and Phytium Arm-based chips to underpin Tianhe-3 and Sunway Oceanlite systems. These efforts support both civilian applications, such as climate modeling and AI research, and military projects, including hypersonic simulation and cyber warfare, demonstrating the dual-use nature of exascale computing.

Meanwhile, the European Union is pursuing digital sovereignty through the EuroHPC Joint Undertaking. Its flagship JUPITER system emphasizes energy-efficient, sustainable HPC, leveraging renewable energy and waste-heat reuse while enabling research in precision medicine, climate science, and AI. Japan, building on the legacy of Fugaku, is advancing successors designed to surpass the exascale threshold, integrating ARM-based CPUs and GPU accelerators optimized for energy efficiency and hybrid AI-HPC workloads.

Other nations, including India, Russia, South Korea, and Saudi Arabia, are also investing in supercomputing infrastructure to ensure strategic and scientific competitiveness. India’s National Supercomputing Mission, for example, aims to develop an indigenous exascale system, aligned with broader goals of technological self-reliance and innovation-led growth.

In this high-stakes global contest, exascale computing is no longer just a measure of raw speed—it is a barometer of geopolitical influence, scientific leadership, and technological resilience. How nations balance performance, sustainability, AI integration, and international collaboration will shape the future of global innovation, defense, and economic competitiveness.

Scientific Revolutions Unleashed

Exascale supercomputers represent a transformative leap in computational capability, with the potential to reshape virtually every domain of science and technology. Fields such as chemistry, materials science, high-energy physics, cosmology, oil exploration, and transportation stand to benefit immensely.

Exascale computing also empowers data-intensive fields like genomics, metagenomics, and microbiome research by processing terabytes of sequencing data in hours rather than days or weeks. These capabilities are central to initiatives like the Human Brain Project, which aims to develop interactive, multi-scale models of the human brain using exascale-class computing. The fusion of big data, machine learning, and high-performance computing is enabling researchers to model complex phenomena—from predicting earthquake impact zones and simulating autonomous vehicle crashes to rendering cinematic visual effects. As computational demands grow exponentially, exascale systems stand at the core of scientific advancement, industrial innovation, and national power in the 21st century.

These machines enable simulations of unprecedented complexity and resolution, allowing researchers to study molecular interactions—such as those between viruses and human cells—with remarkable accuracy. This can accelerate breakthroughs in drug discovery, vaccine development, and the understanding of neurological processes. For example, exascale systems could simulate up to 10% of the human brain’s 100 billion neurons, a major step forward in cognitive and memory research. In materials science, exascale platforms allow scientists to simulate atomic-scale interactions that inform the development of stronger, lighter, and more durable materials for aerospace, construction, and manufacturing.

The healthcare sector stands to benefit immensely from this computational leap. Exascale systems can model molecular interactions and simulate protein folding with atomic-level precision, expediting drug discovery for diseases such as Alzheimer’s, Parkinson’s, and various cancers. What once required years of laboratory testing can now be compressed into days, dramatically reducing time-to-market for life-saving treatments. These advancements also open new frontiers in personalized medicine by enabling the simulation of individual genetic responses to different therapeutic compounds.

Exascale computing is poised to profoundly transform humanity’s understanding of complex systems by offering computational power at a scale never before achieved. In climate science, these machines simulate atmospheric and oceanic patterns with unprecedented granularity, enabling scientists to forecast extreme weather events years in advance. Such predictive capabilities not only aid disaster preparedness but also guide long-term global policy decisions on climate mitigation. For instance, detailed simulations of ocean currents, ice sheet dynamics, and carbon cycle feedback loops are instrumental in assessing the future impacts of sea-level rise and global warming.

On a planetary scale, exascale computing is poised to become indispensable in the fight against climate change. With their immense power, these systems can support models like the Energy Exascale Earth System Model (E3SM), simulating atmospheric, land, ocean, and ice dynamics with unprecedented resolution. This enables far more accurate forecasting of sea-level rise, extreme weather events, and regional climate impacts—critical tools for policymakers and disaster response planning. High-resolution climate modeling requires the ability to run thousands of simulations with slight variations in input to capture chaotic system behavior—something only exascale systems can manage efficiently. As emphasized by European researchers like Dr. Joussaume, maintaining global leadership in climate science demands access to the world’s most powerful computing resources.

The energy sector is another area where exascale computing is set to be a game changer. By analyzing wind flow and atmospheric conditions at fine resolution, these systems can optimize the layout and design of wind turbines, boosting efficiency and lowering costs. Similar advances are expected in solar and nuclear energy, biofuels, and combustion research. Exascale computers also have the potential to vastly improve weather prediction and disaster forecasting by processing more variables and running more simulations in real time. The Energy Exascale Earth System Model (E3SM), for instance, is being developed to simulate Earth’s climate systems with far greater fidelity than today’s models, offering deeper insight into the long-term effects of climate change and helping policymakers and urban planners mitigate future risks.

Exascale level computing could have an impact on almost everything, Argonne National Laboratory Distinguished Fellow Paul Messina said. It can help increase the efficiency of wind farms by determining the best locations and arrangements of turbines, as well as optimizing the design of the turbines themselves. It can also help severe weather forecasters make their models more accurate and could boost research in solar energy, nuclear energy, biofuels and combustion, among many other fields.

In the pursuit of sustainable energy, exascale computing enables the high-fidelity modeling of plasma behavior within fusion reactors—a feat crucial to replicating the Sun’s power on Earth. This capability could eventually unlock a near-limitless, clean energy source, fundamentally reshaping the global energy landscape. Additionally, exascale systems are revolutionizing smart infrastructure by powering digital twins of cities. These real-time, data-driven models facilitate predictive maintenance of critical infrastructure, optimize traffic flows, and support the safe deployment of autonomous vehicles, fostering more resilient and intelligent urban environments.

One of the most promising frontiers is the simulation of complex systems—such as working batteries, fusion reactors, or entire brains—at atomic or molecular scales. In energy storage, for example, exascale computing enables the simulation of complex electrochemical processes inside batteries, helping researchers identify optimal materials and chemical compositions faster and more accurately. These capabilities support the design of next-generation batteries, critical for electric vehicles, renewable energy integration, and portable electronics. For example, scientists can use exascale power to model the behavior of materials inside a battery over time, optimizing chemical compositions for energy density, safety, and longevity.

This brute-force computing capability, however, must be paired with smarter mathematical models and efficient algorithms to yield actionable insights. Researchers at institutions like Argonne National Laboratory and Princeton Plasma Physics Laboratory are already using such capabilities to simulate plasma behavior in fusion reactors, helping advance the feasibility of clean and sustainable fusion energy.

Meanwhile, industries ranging from pharmaceuticals and automotive to logistics and aerospace can leverage exascale-level simulations to design better products, improve safety, and shorten development cycles, reinforcing national economic competitiveness. Together, these applications illustrate that exascale supercomputers are far more than just fast machines—they are engines of innovation that empower humanity to tackle its most pressing scientific, medical, and societal challenges.

Military and Strategic Superiority

Exascale supercomputers have become the linchpin of modern military innovation, underpinning the design, simulation, and deployment of advanced weapon systems that shape global power dynamics. Their immense computational power enables nations to model, optimize, and predict outcomes across domains—from hypersonic weapons to nuclear deterrence and cyber operations.

Hypersonics and the Future of Combat

Exascale systems allow hyper-accurate modeling of hypersonic missiles, where even minor aerodynamic instabilities at Mach 10+ speeds can be resolved in simulations, drastically reducing the need for costly and risky physical testing. Engineers can optimize missile design in real time, ensuring next-generation systems can evade current missile defenses. Both the United States and China are racing to leverage this computational power, recognizing that hypersonic superiority could redefine deterrence and first-strike capabilities.

Beyond missiles, scramjet-powered vehicles could revolutionize space access and global logistics. However, designing scramjets demands simulating turbulent supersonic combustion—an extraordinarily complex challenge. Exascale-enabled multiphysics models integrate aerodynamics, thermal dynamics, materials science, and AI to optimize these systems for stability and heat resistance. Pentagon programs like the Hypersonic Air-breathing Weapon Concept (HAWC) rely on similar high-fidelity simulations, developing Mach 5+ weapons capable of evading advanced air defenses. Exascale simulations also enhance weather forecasting, ensuring that military assets—from fleets to aircraft—operate safely while maintaining strategic positioning.

Intelligence, Cyber Warfare, and the Data Deluge

Modern intelligence and cybersecurity operations are impossible without exascale computing. Agencies such as the NSA and GCHQ process petabytes of data from satellites, drones, and global sensor networks. Exascale enables real-time decryption, AI-driven pattern recognition, and anomaly detection, identifying terrorist networks or geopolitical crises hidden within massive datasets. Programs like DARPA’s Ubiquitous High Performance Computing (UHPC) embed HPC capabilities directly into battlefield systems, giving troops the ability to analyze sensor feeds and predict threats instantly.

In cybersecurity, exascale platforms track malware, map adversarial botnets, and simulate complex cyberattacks, strengthening defense in a domain increasingly defined by speed. The same systems also facilitate the development of quantum-resistant cryptography, creating a perpetual cycle of offense and defense where encryption and decryption compete at machine speed. Losing exascale leadership could compromise missile defense accuracy, where systems like the Glide Phase Interceptor (GPI) rely on processing vast sensor data streams in milliseconds.

Nuclear Deterrence and Virtual Arms Races

Exascale computing also underpins virtual nuclear testing, allowing nations to maintain and modernize stockpiles without detonating weapons. High-resolution simulations model warhead aging, exotic material behavior under extreme conditions, and fusion performance, supporting programs like the U.S. Stockpile Stewardship Program with systems such as El Capitan. These capabilities ensure reliability across the nuclear triad, even as competitors like China and Russia modernize their own arsenals with compact, high-yield warheads and hypersonic glide vehicles.

China’s rumored Tianhe-3 system, reportedly aiding DF-ZF hypersonic glide vehicle development and AI-driven swarm drones, illustrates the risk of asymmetric HPC capabilities. If U.S. exascale systems fall behind, adversaries may field weapons whose capabilities are obscured, forcing strategic miscalculations. Conversely, U.S. dominance via Frontier and El Capitan enables predictive maintenance of assets like the B-21 Raider and optimization of autonomous combat platforms. The Joint All-Domain Command and Control (JADC2) relies on exascale computing to fuse satellite, drone, and battlefield data into a unified “kill web,” a model China’s PLA seeks to emulate through its Strategic Support Force.

Battlefield Simulation and Space Warfare

Exascale computing extends beyond weapons design, enhancing modeling of complex battlefield environments. From large-scale joint-force operations to drone swarm coordination, these machines allow strategists to visualize and refine mission scenarios in virtual theaters. In space operations, exascale systems predict satellite constellation behavior, model space weather disturbances, and optimize orbital asset deployment—crucial for communications, navigation, and space-based defense.

As U.S. General Mark Milley observed, “Whoever masters AI and quantum computing will rule the battlefield.” At the heart of this emerging dominance lies exascale computing—the digital backbone of 21st-century strategic power.

Conclusion: Securing the Exascale Edge

Exascale computing is redefining national security paradigms. From securing nuclear arsenals to decrypting massive intelligence datasets, these machines are central to strategic decision-making. Yet challenges remain, including AI-driven adversarial threats, supply chain vulnerabilities, and energy-intensive operation. The next frontier—combining exascale systems with quantum accelerators and neuromorphic architectures—will determine whether democracies or autocracies dominate future warfare. In this era, exascale is more than a technological milestone—it is the cornerstone of geopolitical survival.

Challenges: The Roadblocks to Exascale Dominance

While exascale computing promises transformative benefits, achieving and sustaining it comes with formidable challenges—foremost among them is power consumption. A single exascale system can draw between 20 to 60 megawatts of electricity, comparable to the energy needs of a small city. This immense demand has driven a global push toward energy-efficient architectures and next-generation cooling solutions. Techniques such as liquid immersion cooling, direct-to-chip thermal management, and advanced airflow systems are now being deployed to prevent overheating and reduce energy waste. Researchers are also exploring disruptive technologies—including neuromorphic computing, quantum processors, and optical interconnects—to push performance higher without a corresponding surge in energy use.

Beyond hardware, software remains a major bottleneck. Much of today’s scientific and engineering code was designed for older HPC generations and cannot fully exploit the massive parallelism of exascale architectures. Scaling applications to run efficiently across millions of processing cores requires rewriting legacy software and developing entirely new algorithms that dynamically balance workloads, optimize memory usage, and maintain fault tolerance. This represents a profound shift in computational paradigms, demanding new skills, programming models, and a rethink of long-established scientific workflows.

Geopolitical tensions add yet another layer of complexity. The ongoing U.S.-China technology rivalry—particularly around advanced semiconductor supply chains—has resulted in export restrictions and chip bans, threatening to fragment global HPC collaboration. In response, nations like China are investing heavily in self-reliant chip production and domestic high-performance computing ecosystems, aiming to insulate themselves from foreign dependencies. This push for technological sovereignty risks creating parallel HPC ecosystems, potentially slowing the pace of global scientific progress in a field that thrives on collaboration.

Despite these obstacles, the global pursuit of exascale computing continues unabated. The stakes are too high—from modeling complex climate systems and accelerating drug discovery to maintaining strategic military advantages. Overcoming challenges in energy efficiency, software scalability, and geopolitical friction will be as critical as achieving raw computational power, defining not just the future of supercomputing, but the global balance of scientific and strategic capabilities.

The Future: Beyond Exascale and Into the Quantum Era

Even as the world celebrates the exascale milestone, the race for computational supremacy is evolving at a staggering pace. Nations are already looking beyond exascale to zettascale computing—systems capable of performing one sextillion (10²¹) calculations per second, a thousandfold leap beyond today’s fastest supercomputers. These machines promise real-time, planet-wide simulations, enabling detailed climate modeling, rapid disaster response, and advanced materials design on scales previously unimaginable.

At the same time, countries are investing in neuromorphic computing, architectures inspired by the human brain. These systems mimic cognitive processes like perception, learning, and reasoning, opening the door to breakthroughs in pattern recognition, autonomous decision-making, and AI-driven scientific discovery. Imagine machines capable of instantaneous language translation across all dialects, or AI systems that predict global supply chain disruptions before they occur. Together, zettascale and neuromorphic technologies could redefine the boundaries of science, society, and human-machine interaction.

Yet, with these extraordinary capabilities come profound ethical and geopolitical dilemmas. Will this new computational power accelerate autonomous weapons development and cyber warfare, intensifying global tensions? Or can it foster unprecedented collaboration in confronting existential challenges such as pandemics, climate change, and resource scarcity? Efforts like the U.S.-EU Trade and Technology Council are seeking shared governance frameworks, but achieving transparency and trust in a world of ultra-powerful computational assets remains a critical challenge.

Quantum on the Horizon

While exascale represents the peak of classical computing, quantum computing is emerging as the next transformative frontier. Leveraging qubits and principles of superposition and entanglement, quantum systems promise solutions to problems beyond the reach of even the largest exascale machines. Applications span unbreakable encryption, revolutionary drug discovery, and optimization challenges across logistics, finance, and climate modeling.

However, today’s quantum computers face significant technical hurdles, from qubit instability and short coherence times to the need for robust error-correction protocols. As a result, exascale systems will remain the backbone of high-performance computing over the next decade, running massive simulations, powering AI models, and eventually serving as platforms to integrate quantum accelerators. The future of supercomputing is likely hybrid, combining the raw brute force of exascale with the problem-solving potential of quantum processors, creating unprecedented computational synergy.

A New Era of Power Dynamics

Exascale supercomputers are far more than instruments of calculation—they are strategic assets that will shape the contours of global influence in the 21st century. Leadership in exascale computing has become a proxy for dominance in science, defense, and economic innovation. Those who command this power will define the future of AI, climate modeling, energy security, and next-generation defense systems.

The race mirrors the space race in its ambition and stakes, offering prestige, scientific leadership, economic spinoffs, and strategic advantage. From simulating nuclear stockpiles to predicting extreme weather, exascale machines will push the limits of what humanity can understand and achieve.

But this power brings immense responsibility. The exascale race cannot be allowed to devolve into a zero-sum contest of military escalation or technological isolation. Instead, it presents an opportunity to establish new models of international collaboration, ethical standards, and shared technological stewardship. Innovation must be guided by foresight; progress must be anchored in principles.

Ultimately, the race to exascale—and beyond—will determine who steers the future of humanity. The choices made today by scientists, policymakers, and global leaders will decide whether this extraordinary computational power serves the global good or entrenches division and strategic imbalance. In this new epoch, exascale is not just a technological milestone—it is the cornerstone of 21st-century influence and human progress.