Juicing the Nuclear Weapons Complex

Marty Pfeiffer (@nuclearanthro.bsky.social on Bluesky) is a gadfly to DOE and NNSA, mainly through their Freedom of Information Act (FOIA) programs. He has been complaining that NNSA has not published its annual mandated report to Congress on Stockpile Stewardship, the program that monitors and researches the operation of nuclear weapons.

This week he found out why. He turned up another document.

To be fair, the Los Alamos Study Group found the document first and had it on their website. It’s a memorandum from David E. Beck, Deputy Adminstrator for Defense Programs, National Nuclear Security Administration, dated February 11, 2026.

The memo presents a framework, “Responsive Today, Dominant Tomorrow: Enhancing American Nuclear Dominance.” The managers of the various components of the nuclear weapons complex are ordered to prepare implementation plans. I can imagine it’s been a bad time back at my old employer, Los Alamos National Laboratory.

“These implementation plans should include detailed sub-objectives, tasks, and milestones that will support each objective.” The implementation plans were due on March 7.

The framework seems to include what is known as Stockpile Stewardship as well as nuclear weapons manufacturing, presumably the reason that the Stockpile Stewardship report has not emerged. Stockpile Stewardship has been the program of investigating nuclear weapons behavior through computing and basic science.

The bottom line message of the memo seems to be MOAR NUKES, although apparently without any increase in funding or personnel. And it’s supposed to happen by the end of calendar year 2028.

So many things I might address in the memo. Let me start with an overview.

This memo belongs to a genre of promises being made to Donald Trump by defense-related officials appointed by him. “Golden Dome” is one. Trump’s battleships are another. Trump wants the biggest, baddest military ever. The US military has been, at least, the biggest for some long time. Drone warfare is drawing questions about baddest. But Trump sees it in terms of the most and biggest. MOAR is BETTER!

In his first term, he expressed a desire to go back to the numbers of nuclear weapons at the maximum of the Cold War. That would be around 30,000 nuclear weapons in contrast to the 3700 the US has today. It would also be ridiculous. In any case, it’s not possible. But Beck wants to make a run at it.

Why would it be ridiculous? Because Russia has around 4400 and having roughly equal stockpiles and deployed weapons has long been a part of deterrence. Because China may be building up, but they currently have many fewer at around 600. But what if both decided to gang up on us? Well then all bets are off, and detonating thousands of nuclear weapons is the worst idea of all. Detonating even a few of them is a bad idea too.

The Russians and Chinese know this.

The United States has limited ability to build new nuclear weapons. The only facility that can build the plutonium pits essential to nuclear weapons is Los Alamos. Another is being built at Savannah River, to be ready in 2030. Over the course of the last year or so, Los Alamos has turned out something like ten pits. The Beck memo states a goal of 60 pits a year by 2028. An earlier goal of 30 pits a year by 2030 was deemed impossible by the previous NNSA administrator.

The facility that built those Cold War thousands of pits was the Rocky Flats plant in Colorado, with plutonium fires and environmental damage along the way. It was closed in 1992.

Beck’s massive ambitions are matched with an absence of increased funding. Perhaps some of the $1.5 trillion being floated for defense putposes is intended to go to the nuclear weapons complex, but historically the two budgets have been separate.

The memo also wants AI to be inserted into various processes. This may make sense in some places, but it’s clear that Beck and his people haven’t figured out which places those are. It’s a statement of faith in AI derived from the DOGE boys.

Two issues read between the lines: pressing the weapons laboratories (Los Alamos, Livermore, and Sandia) toward a single-minded weapons mission. All three of those laboratories have significant portions of their effort in basic research across a range of topics, including human health, space science, and civilian energy sources and usage. Historically, they have looked on that diversity (uh-oh, bad word!) as contributing to a broad base from which weapons science can draw. They will not want to give that up, although perhaps the gutting of US science can force them to.

The second issue is an apparent desire by Trump and other people in his administration to return to nuclear explosive testing. That objective is not expressed specifically in the Beck memo, but some of the language suggests it.

Robert Peters of the Heritage Foundation, for example, argues that the purpose of nuclear explosive testing would be

successfully deterring a strategic attack on the United States or one of its closest allies by a nuclear-armed adversary. And if an American President deems it necessary to conduct a nuclear explosive test to convince an adversary that it cannot escalate its way out of a conflict, the U.S. nuclear enterprise should be ready to respond.

This is a particularly silly argument. The Russians know that the US has nuclear weapons. The Chinese know that the US has nuclear weapons. What would a nuclear explosive test add?

Another reason that might be adduced is that questions may exist about the readiness of nuclear weapons as they age that cannot be answered by Stockpile Stewardship. Over the past several years, there have been rumors of concerns by the JASONs advisory group that pits may need to be remanufactured. This is not mentioned in the Beck memo.

In any case, diagnostic nuclear explosive testing will not take place any time in the near future. The Nevada Test Site (now the Nevada National Security Site, significantly) is no longer suitable for such tests. Las Vegas has grown much too close. A new site would be needed. And once the new site has been acquired and prepared, advances in instrumentation would make it necessary to reinvent the diagnostics. An atmospheric test without diagnostics, purely to illustrate that we have nuclear weapons, could be done if the objections from potential downwinders were ignored. And what if that particular weapon turned out to be a dud? How much deterrence would that add?

Peters’s argument suggests that the underlying reason for wanting nuclear explosive tests is to prove something to themselves, as does the dominant dominance theme of the memo. In a similar vein, it looks like Stockpile Stewardship may be renamed Deterrence Science, so Marty is going to have to change his alerts. His article covers the Beck memo differently than I’ve done and is worth reading.

The implementation plans have been submitted or extensions set. It would be interesting to know whether the managers of the nuclear weapons complex have provided an honest evaluation or supported the wishful thinking. It looks like the people of Golden Dome and the battleship project have gone for the latter.

Cross-posted to Nuclear Diner