The U.S. has spent the last decade buying rare earth magnets from China at commodity prices and calling it a solved problem. On June 19, ARPA-E announced it is not solved, not even close. The agency deployed $72 million across two programs, ROCKS and MAGNITO, with a blunt mission: compress mineral exploration from a decade to 1-2 years, and build permanent magnets twice as strong as anything in production today. The announcement comes with a quote from ARPA-E Director Conner Prochaska that does not hedge: "Rare earth elements, critical minerals, and magnets are indispensable to American energy, industry, and national security." This is government-speak for: we have been exposed, and we are moving.
The $72 million is part of a broader $100 million ARPA-E initiative announced in 2025 to secure domestic critical minerals. ROCKS targets the discovery side, the program aims to develop drilling, sensing, and analytical technologies that cut the timeline for assessing whether a mineral deposit is actually worth mining. Currently, a company spends years just figuring out whether a deposit is economically viable; ROCKS wants to do that in months. MAGNITO targets the manufacturing side, it aims to discover materials with magnetic saturation moments in excess of 2 to 2.5 Tesla, roughly double the strength of the best permanent magnets in commercial production today. South Dakota School of Mines received $3.1 million and the University of Houston received $2.8 million as initial awardees.
Why magnet strength matters: every electric motor, in an EV, a wind turbine, an industrial pump, a compressor, uses permanent magnets. A magnet twice as strong means you need half as much rare earth material per motor, and you can shrink the motor itself. For EVs, that cuts weight and cost. For wind turbines, it reduces materials cost in the nacelle by 20-30 percent. For industrial motors, which consume half of all U.S. industrial electricity, efficiency gains compound across millions of units. The physics is not theoretical; it is the single largest lever for reducing rare earth consumption per watt of output. ARPA-E is betting that American labs can crack materials that were previously thought impossible using computational tools and high-throughput experimental techniques, specifically by exploring alloys made of three, four, or five elements combined in ways that have never been systematized before.
Here is the hard part: ROCKS and MAGNITO are research programs, not supply chain programs. South Dakota and Houston will publish papers, not tons of rare earth ore. The real test comes after 2027, when ARPA-E's funding cycle completes and the technologies face actual mining and manufacturing scale-up. That is where the story either becomes real or becomes another government R&D program that succeeded in the lab. Domestic rare earth mining in the U.S. has been uneconomical for thirty years, not because deposits do not exist, but because Chinese production is subsidized enough to undercut any domestic competitor. A faster exploration timeline and a better magnet help, but they do not solve the subsidy problem. ARPA-E has essentially bet that technology acceleration can overcome a policy problem.
Who wins: American EV and wind manufacturers see reduced long-term magnet costs and supply certainty, though not for 3-4 years minimum. Defense contractors building hypersonic systems and precision radar gain access to stronger magnets, improving system performance. Who does not win: Chinese rare earth processors lose some of the leverage they currently hold over U.S. supply chains, but only if ROCKS succeeds in commercialization. Chinese rare earth miners and magnet manufacturers lose pricing power as American supply ramps, which is why watch for Beijing to restrict rare earth exports or rare earth element processing in the next 12 months, a move that could actually accelerate American domestic investment even further.
Two specific markers determine whether this program delivers real supply chain impact. First, by Q4 2026, watch for project selections under RECOVER, the parallel program managing critical mineral recovery from domestic wastewaters (including municipal, agricultural, mining, and industrial wastewater streams), if ROCKS and RECOVER move in coordination, with shared timelines and integrated technical goals, the U.S. signals intent to build a closed-loop system. Second, watch for the first commercial mining operations using ROCKS-derived technologies; even a pilot project at a domestic deposit by late 2027 would indicate the ten-year exploration timeline is actually compressing. If neither happens by then, this is just expensive R&D.