Ideas
Hyperscaler PPAs unlock nuclear renaissance.
Hyperscalers and data center builders are now willing to sign 20-year $100 per megawatt hour PPAs, creating order books and secure offtake that make advanced nuclear projects financable and open the nuclear renaissance.
Restart old nuclear plants fastest.
Uranium conversion capacity is highly concentrated, with much capacity in Russia and only one old US plant; executive orders require exiting Russian conversion supply by 2028 but no alternative yet exists, so companies filling the gap are well positioned.
Mass-produced SMRs cut nuclear costs.
Mass manufacturing reactor components and small shippable reactor units can shrink lead times from years to months and drive nuclear costs down to compete with diesel first and later other generation; reference plants by 2028 or 2029 start this cost curve.
Microreactors replace remote diesel generators.
Advanced microreactors can replace large diesel generator sets at remote mining and Arctic installations, eliminating weekly fuel flights and creating large cost savings even though that niche is smaller than data centers.
US needs exportable reactor model.
Global nuclear construction is split between Chinese and Russian designs, while the US lacks a proven on-time, on-budget reactor model; getting a US reference microreactor built creates a pathway to larger exportable US reactors.
Laser enrichment may dominate uranium enrichment.
Laser enrichment can selectively excite U-235 rather than both isotopes, making enrichment much more efficient than gas centrifuges; GLE and Hexium are pursuing laser and lithium routes, so laser enrichment is likely the future though not fully proven.
Supercritical CO2 turbines boost efficiency.
Supercritical CO2 power cycles allow higher pressures, much smaller turbine equipment, mass manufacturing, and up to 50% thermal efficiency; bottoming cycles can add another 10% to 20% more power generation from existing assets.
Fusion could undercut fission by 2030s.
Fusion power plants could be operational in the early to mid 2030s, around the same time as larger SMRs, and should cost less than fission because hydrogen fuel eliminates uranium mining, conversion, enrichment, and storage; HTS magnets and lasers have dramatically shrunk devices.
Fusion enabling tech and superconducting magnets.
Both integrated fusion power plant developers and enabling component suppliers are attractive; high-temperature superconducting magnets and laser technologies are key, and Spinner Energy applies superconducting magnets to data center and grid power transfer.
Silicon carbide enables compact power electronics.
Silicon carbide became cheap thanks to electric vehicle scale and enables small, high-voltage, thermally tolerant power devices; it is important for solid-state transformers, grid interconnections, and nuclear cladding, with wafering advances like Halo Industries.
Solid-state transformers cut interconnection lead times.
Transformers are a limiting grid interconnection component; companies are mass manufacturing solid-state transformers using silicon carbide to cut lead times from years to weeks or months, which is essential to connect more nuclear and data center power.
Rare earths and magnets attractive.
Upstream material production for the electric grid and energy hardware, including rare earth element refining, mining, and magnet production, is an attractive opportunity as new energy and grid equipment scales.
Copper and aluminum supply chains tight.
Copper and aluminum are critical for the electric grid and manufactured components; US aluminum refining largely disappeared and new mine timelines are scary, so shrinking timelines and adding supply is a key opportunity.
800V data centers need power management.
Nvidia's 800V data center distribution roadmap and the scale of one megawatt per cabinet create a need for low-cost, standardized rack-level power management and smooth grid interfaces because AI training loads swing violently.
Virtual power plants reduce peak load.
Aggregating flexible demand into virtual power plants can reduce peak load and behave like a real power plant, reducing the need to build additional generation capacity.
Geothermal depends on deep drilling advances.
Geothermal is attracting more activity and could reach 5 to 7 cents per kilowatt hour if deep drilling and power density improve; Quaise is working on cost-effective deep drilling near load, and drill-bit materials are a cross-industry opportunity.
Industrial robotics dominate near term.
Industrial robotics in dangerous, remote, or difficult tasks is seeing the most penetration now, including crop dusting via Pyka, shipbuilding, decommissioning, and construction, while home humanoid robots are five-plus years out; robotic safety remains key.
Industrial robotics dominate near term.
Industrial robotics in dangerous, remote, or difficult tasks is seeing the most penetration now, including crop dusting via Pyka, shipbuilding, decommissioning, and construction, while home humanoid robots are five-plus years out; robotic safety remains key.
This Macro Voices video, published September 03, 2026,
features Carly Anderson, Erik Townsend
discussing Advanced nuclear energy, URA, Small modular reactors, Microreactors, NLR, Uranium enrichment, Supercritical CO2 turbines, Nuclear fusion, Fusion enabling technologies, High-temperature superconducting magnets, Silicon carbide, Wide bandgap semiconductors, Solid-state transformers, REMX, COPPER, Aluminum, Data center power management, Virtual power plants, GEOTHERMAL ENERGY, Industrial robotics, Humanoid robotics.
18 trade ideas extracted by AI with direction and confidence scoring.
Speakers:
Carly Anderson,
Erik Townsend
· Tickers:
Advanced nuclear energy,
URA,
Small modular reactors,
Microreactors,
NLR,
Uranium enrichment,
Supercritical CO2 turbines,
Nuclear fusion,
Fusion enabling technologies,
High-temperature superconducting magnets,
Silicon carbide,
Wide bandgap semiconductors,
Solid-state transformers,
REMX,
COPPER,
Aluminum,
Data center power management,
Virtual power plants,
GEOTHERMAL ENERGY,
Industrial robotics,
Humanoid robotics
