Some old websites survived because nobody touched them. The Nuclear Weapon Archive survived because people kept having to move the whole thing somewhere else.
The Nuclear Weapon Archive is still online.
Its own history page traces the project back to November 1994, when Gary Au, then a student at Melbourne University in Australia, founded the High Energy Weapons Archive. That puts the site’s ancestry inside the first commercial-browser era of the Web, when a serious reference project could still begin as a directory on a university machine and become internationally known before anyone had settled on what a “web publication” was supposed to look like.
The archive did not enjoy a quiet childhood. Traffic surged around the fiftieth anniversary of the atomic bombing of Japan in 1995, drawing enough attention that Melbourne University shut the site down under its computing-use rules. Au later took a job with the Australian Department of Defence and handed administration to Carey Sublette in January 1996. From there the material bounced among hosts in Finland, Italy, and the United States, including the Federation of American Scientists and Enviroweb.
That history explains one of the site’s most important design decisions. In 2003 it finally acquired nuclearweaponarchive.org, specifically to stop years of host changes from breaking every incoming link. The old High Energy Weapons Archive name also became The Nuclear Weapon Archive.
What survives is enormous. There are historical sections on the Manhattan Project, Trinity, Hiroshima and Nagasaki, nuclear test programs, national arsenals, weapon-development history, reference material, photographs, timelines, and a long FAQ. The site also records its own update history instead of pretending it appeared fully formed yesterday. Entries jump from the early 2000s through later maintenance work, including updates in 2020 and 2023.
The important thing in 2026 is not that every page is visually ancient. It is that this is an example of a serious specialist reference site carrying its institutional memory with it. Modern sites are often rebuilt until their own history disappears. This one tells you who started it, who inherited it, why it moved, why the domain exists, and what each phase was trying to preserve.
That makes the archive itself part of the historical record, separate from the difficult subject matter it documents.
If you want to see how a Web reference project can survive repeated displacement without turning into a content-platform blob, go through the archive itself. CacheRat’s larger 1,967 Ancient Web Domains research list contains more places built with the same old assumption: if the information matters, put it on a page and keep the page alive.