Note that there are two main types of DOE labs: NNSA (Sandia, Los Alamos, Livermore) and Office of Science (Brookhaven, Berkeley, Oak Ridge, Argonne, ...). Although the former is more focused on "nuclear weapons stockpile stewardship", there is still much basic science at all DOE labs, especially where computer science meets physics and other domain sciences.
Perhaps relevant to HN, I would mention the Applied Computer Science group at Los Alamos, which is in hiring mode (https://www.lanl.gov/org/ddste/aldsc/computer-computational-...). Besides supporting computational physicists in code development efforts, this group does a variety of researchy things like designing programming model, doing compiler development, building ML models, especially with an eye towards large scale scientific computing. The pay at a DOE lab is less than FAANG (PhD student interns might be around $80k/yr and starting staff scientists maybe $130k/yr), but the tradeoff for some people would be the research-flavor of the work, and the flexibility. Many of the LANL codes being developed are open source, for example. Other DOE labs have similar computer science divisions. For example, Oak Ridge, Argonne, and Berkeley all have "leadership computing" facilities.