America’s Nuclear Secret Weapon Is Not a New Bomb. It Is the Science Keeping Old Ones Credible

As Russia expands its arsenal, China builds new warheads and North Korea rejects denuclearization, the United States is relying on an extraordinary network of laboratories, supercomputers and underground experiments to maintain its deterrent without explosive nuclear testing.

By David Soyer Follow

WASHINGTON | Published at 1:51 p.m. ET

Somewhere beneath the Nevada desert, American scientists are conducting experiments that create enormous pressure without producing a self-sustaining nuclear chain reaction. In California, one of the world’s most powerful computers is running detailed simulations of materials under conditions that cannot be safely reproduced in an ordinary laboratory. At facilities in New Mexico, technicians are rebuilding the industrial capability to manufacture plutonium components that sit at the center of nuclear weapons.

None of these projects has the immediate drama of a missile launch or the visible menace of a mushroom cloud. Together, however, they form what may be America’s most important advantage in the renewed global nuclear competition.

The United States’ secret weapon is not a mysterious warhead hidden from public view. It is Stockpile Stewardship, the vast scientific and industrial system used to determine whether nuclear weapons will remain safe, secure and effective without conducting a full explosive nuclear test.

That distinction matters as the international security environment grows more dangerous.

The New START arms-control treaty between the United States and Russia expired in February 2026 without a replacement, ending the last major agreement placing numerical limits on the two countries’ strategic nuclear arsenals. China continues to expand its nuclear force. North Korea said Thursday that its nuclear status is irreversible and pledged to keep strengthening its arsenal. At the same time, artificial intelligence, cyberwarfare, hypersonic weapons and advanced missile defenses are complicating calculations that once rested largely on the balance between American and Soviet forces.

The nuclear competition is no longer a two-country race measured only by the number of warheads.

It is now a contest over whose weapons can survive, communicate, penetrate defenses, resist cyberattack and remain reliable after decades in storage. America’s advantage lies in its ability to combine experimental physics, advanced computing, precision manufacturing and decades of weapons data.

That advantage is real, but it is not guaranteed.

The arsenal is old, even when the components are new

The United States has not conducted an explosive nuclear test since September 1992. Its last test, code-named Divider, was carried out underground at the Nevada Test Site.

Since then, the country has observed a testing moratorium while continuing to certify its nuclear stockpile.

The weapons in the current arsenal are based on designs developed and tested during the Cold War. Many have received replacement components, upgraded safety systems and life-extension work, but the underlying designs may be several decades old.

A nuclear weapon is not simply a sealed metal object placed in storage. Its metals, high explosives, polymers, electronics and other materials change over time. Some components corrode. Certain materials become brittle. Radioactive decay affects the plutonium at the weapon’s core. Replacement parts may not behave exactly like components manufactured with materials or processes that are no longer available.

Those changes create a difficult national-security question: How can the government be certain that a weapon will work as intended without detonating one?

Stockpile Stewardship is the American answer.

The program combines historical test information with laboratory experiments, surveillance of deployed weapons, component disassembly, advanced imaging and computer simulations. Scientists use the results to understand how an aging weapon is changing and whether it can still perform within its certified requirements.

The National Nuclear Security Administration, a semiautonomous part of the Department of Energy, is responsible for the program.

This is the foundation beneath the more visible nuclear triad of land-based missiles, ballistic-missile submarines and strategic bombers.

Missiles and aircraft can deliver a weapon. Submarines can hide it. The stewardship system is what gives military planners confidence that the warhead itself remains dependable.

Testing without creating a nuclear explosion

One of the program’s most important tools is the subcritical experiment.

During such an experiment, scientists subject plutonium or other materials to high pressure and extreme conditions. The quantity and arrangement of nuclear material remain below the threshold needed for a self-sustaining chain reaction, so the test does not produce a nuclear yield.

The experiments allow researchers to measure how plutonium behaves as it is compressed, heated and shocked. That information is fed into computer models used to evaluate weapon performance.

The National Nuclear Security Administration conducted its 34th subcritical experiment in May 2024 at the Nevada National Security Site. The experiment, known as Nimble, was the first in the agency’s new Cygnus commissioning series, according to the Arms Control Association.

The experiments take place in facilities designed to contain hazardous materials and gather data at extremely high speed. The Nevada National Security Site says its underground laboratory supports the Stockpile Stewardship Program and helps verify the safety and reliability of the nuclear arsenal.

These are not weapons tests in the traditional Cold War sense. They do not prove an entire warhead design through a nuclear detonation. Instead, they isolate specific physical questions and produce data that can improve the computer models used for certification.

That is both the strength and limitation of the method.

The United States can learn a great deal without breaking its testing moratorium, but its confidence depends on the accuracy of models, the quality of experimental data and the judgment of scientists who understand the original weapons.

The supercomputer at the center of the race

The computational part of this system is advancing rapidly.

El Capitan, installed at Lawrence Livermore National Laboratory in California, was built primarily to support nuclear stockpile stewardship. It became the world’s fastest publicly ranked supercomputer in November 2024, exceeding two quintillion calculations per second in a standard performance test.

The machine gives scientists the ability to conduct simulations with greater resolution and complexity than earlier systems. Researchers can model the behavior of materials, radiation and hydrodynamic processes across scales that would be impossible to calculate manually.

That does not mean a computer can perfectly reproduce a nuclear explosion.

Simulations depend on equations, assumptions and experimental data. A faster machine can calculate a flawed model more quickly, so computing power alone is not enough. The model must be tested against information from historical nuclear detonations and modern laboratory experiments.

America’s advantage comes from possessing all three elements:

Data from more than 1,000 nuclear tests conducted before the 1992 moratorium.
Modern experimental facilities capable of measuring materials under extreme conditions.
Exascale computing capable of combining that information into highly detailed simulations.

No single element is the secret weapon. The integrated system is.

Countries developing entirely new designs without a comparable history of explosive testing may have fewer data points available for validating their models. Countries with extensive historical information but aging scientific infrastructure may struggle to make full use of it.

The United States has both an enormous test archive and an unmatched network of national laboratories. That combination is difficult to replicate quickly.

The National Ignition Facility has a defense mission too

Lawrence Livermore’s National Ignition Facility is widely known for achieving fusion ignition, a scientific milestone in which a fusion reaction produced more energy than the laser energy delivered to its fuel target.

The facility is frequently discussed as a possible step toward clean fusion energy. Its more immediate mission, however, includes nuclear weapons research.

NIF’s 192 laser beams can deliver energy to a tiny target, creating temperatures and pressures that help scientists study processes relevant to nuclear weapon performance. The experiments do not detonate a bomb, but they provide information about fusion, radiation transport and materials under extreme conditions.

This dual role is a key feature of America’s nuclear-science system.

Facilities built for stockpile stewardship can produce discoveries relevant to energy, astrophysics and materials science. Civilian research, in turn, develops instruments and expertise that may support national-security work.

That is why America’s university system, national laboratories and technology companies are part of the strategic balance even when their work has no visible connection to a missile silo.

A race that now involves China

Russia still possesses the world’s largest or one of the world’s largest nuclear arsenals, depending on how different categories of warheads are counted. It also inherited an enormous testing and weapons-development infrastructure from the Soviet Union.

China presents a different challenge.

The Pentagon has repeatedly said Beijing is expanding its arsenal more quickly than previous American projections anticipated. China has built new missile-silo fields, improved its ballistic-missile submarines and developed a broader range of delivery systems.

In February 2026, the United States accused China of conducting a secret low-yield nuclear test in 2020. China denied the allegation, and international monitoring did not produce publicly conclusive proof of an explosive test. Reuters reported that the accusation arrived as Washington sought a new arms-control structure involving both Russia and China.

The dispute reveals the central strategic problem.

If one country believes another is secretly gathering explosive test data, political pressure could grow for the United States to resume testing. Supporters could argue that a new test is necessary to validate modern designs or demonstrate resolve. Opponents could warn that resuming explosive testing would encourage Russia, China, India, Pakistan or North Korea to conduct their own tests.

The United States maintains the Nevada site in a condition that could support a return to testing if ordered by the president. Yet such a decision would carry diplomatic, environmental and strategic consequences.

Stockpile Stewardship gives Washington another option. It allows the United States to maintain confidence in its weapons while preserving the moratorium.

That may be one of the system’s greatest values. It is not merely a technical program. It is also an arms-control tool.

North Korea’s message makes the competition more urgent

North Korean Vice Defense Minister Kim Kang Il said at the Beijing Xiangshan Forum on Thursday that Pyongyang would continue expanding its nuclear force. He described the country’s status as a nuclear-armed state as irreversible.

The comments were reported by Reuters amid continuing tensions among North Korea, South Korea, Japan and the United States.

North Korea’s arsenal is much smaller than those of the United States or Russia, but size is not the only measure of risk. Its expanding missile capabilities, uncertain doctrine and proximity to major population centers create an acute regional danger.

The United States therefore has to maintain extended deterrence, meaning it promises to help defend allies such as South Korea and Japan using the full range of American capabilities.

That commitment must be credible. Allies need confidence that American forces will remain reliable. Adversaries need to believe that an attack would carry unacceptable consequences.

The scientific credibility of the stockpile supports that political credibility.

The weakest link may be manufacturing

America’s laboratory system is formidable, but the nuclear enterprise faces serious industrial problems.

Many production facilities are old. Skilled workers are difficult to replace. Supply chains include specialized materials and components that cannot be purchased from ordinary commercial vendors. Projects frequently run over budget or behind schedule.

Plutonium-pit production is one of the most prominent challenges.

The pit is the plutonium core that initiates the primary stage of many nuclear weapons. The United States lost the ability to manufacture pits at sustained Cold War rates after the closure of the Rocky Flats Plant in Colorado.

The government has been attempting to establish production at Los Alamos National Laboratory in New Mexico and the Savannah River Site in South Carolina. The stated objective has been to produce at least 80 pits per year across the two locations, but government reviews have repeatedly warned that the schedule is difficult and costs are rising.

This is where the “secret weapon” description meets reality.

Scientific understanding cannot substitute indefinitely for manufacturing capacity. A simulation can identify a component that needs replacement, but factories and skilled technicians must still build it.

The United States may possess the world’s most sophisticated nuclear-weapons science while struggling to manufacture certain components at the desired speed.

That gap is one of the central risks in the modernization program.

The visible triad is also being rebuilt

While the laboratories preserve and modify warheads, the Defense Department is replacing the systems that carry them.

The Columbia-class submarine is scheduled to replace the aging Ohio-class ballistic-missile fleet. The B-21 Raider bomber is intended to become a central part of the future airborne deterrent. The Sentinel intercontinental ballistic missile program is supposed to replace the Minuteman III.

All three programs are expensive and technically demanding.

Sentinel has faced particularly serious cost growth and schedule concerns. The program involves much more than a new missile. It requires changes to command-and-control systems, launch facilities, communications networks and hundreds of sites across several states.

These delivery systems receive most of the public attention because they can be photographed, counted and compared.

Yet a new submarine carrying an unreliable warhead would not provide a credible deterrent. A modern missile connected to an insecure command network would also be inadequate.

The true system includes the delivery vehicle, the warhead, communications, warning satellites, command authorities, cybersecurity and the laboratories that certify the weapon.

Artificial intelligence introduces a new risk

The nuclear competition is also colliding with the rise of artificial intelligence.

American and Chinese security experts have proposed safeguards to keep AI from making autonomous decisions involving nuclear weapons. Recommendations reported Thursday by Reuters include maintaining meaningful human control and establishing channels for managing AI-related military incidents.

AI can help analyze sensor information, detect unusual activity and strengthen cybersecurity. It may also introduce false confidence, accelerate decisions or generate convincing but incorrect assessments during a crisis.

In nuclear command and control, speed is not always an advantage.

A system that reacts faster than political leaders can understand may increase the risk of accidental escalation. America’s scientific advantage will matter only if it remains paired with disciplined human judgment.

The nuclear race is ultimately a competition in confidence

Nuclear deterrence is built around belief.

American leaders must believe the weapons are reliable. Military personnel must believe their orders and communications are authentic. Allies must believe American security guarantees are credible. Adversaries must believe that aggression would fail or impose intolerable costs.

Stockpile Stewardship helps construct that confidence without requiring the United States to detonate a nuclear weapon.

Its national laboratories are not secret. El Capitan’s existence is public. The National Ignition Facility hosts scientific research. Nevada’s subcritical experiments are announced after they occur.

What remains hidden are the classified details linking those capabilities to specific warheads and military requirements.

America’s “secret weapon” is therefore hiding in plain sight.

It is the ability to turn decades of test data, modern experiments, exascale computing and specialized engineering into a technical judgment that the arsenal will work without proving that judgment through a new nuclear explosion.

That system gives the United States an extraordinary advantage in the nuclear race. It also imposes an extraordinary responsibility.

If the laboratories lose expertise, if manufacturing programs fail, if modernization costs overwhelm budgets or if political leaders treat nuclear weapons as ordinary instruments of war, the advantage could erode.

The most powerful part of America’s deterrent may not be the explosive force inside its warheads. It may be the scientific discipline that has allowed the country to avoid detonating one for more than three decades.

Reporting and interview disclosure

This article is an original Consumerlite News analysis based on public government materials, laboratory documentation, arms-control records and current international reporting.