Nobody said that commercializing fusion power would be cheap or quick.
TAE Technologies said this week it raised another $150 million in a funding round that included investments from existing backers Google, Chevron, and New Enterprise associates.
By the nearly 30-year-old company’s counting, it’s TAE’s twelfth round of investment. To date, it has raised about $1.8 billion, according to PitchBook, making it one of the highest funded fusion companies.
TAE, formerly known as Tri Alpha Energy, worked for years in stealth developing its reactor design. The company initially used a process that started by firing two plasma balls at each other and then spinning the resulting blob with particle beams. The plasma blob — which looks like a hollow cigar — generates its own magnetic field, working alongside the reactors magnets to keep the plasma contained.
In April, the company announced it no longer needed to fire two plasma balls to kick off a reaction. Instead, it was able to form a plasma, heat it, and stabilize it using the particle beams alone. Eliminating that equipment makes the reactor smaller, cheaper, and easier to operate, TAE said.
Google has participated in two rounds of investment in TAE; the previous $250 million round closed in 2022. The tech company has been working with TAE for longer. Since 2014, Google computer scientists have worked with engineers at TAE to use machine learning (a form of AI) to find the ideal settings for a fusion device.
Before AI, the optimization process used to take two months — “about 1,000 experiments,” TAE CEO Michl Binderbauer told me in 2022. AI cut that down significantly, reducing the number of experiments by two orders of magnitude, which could be completed in a few hours.
Today, TAE’s reactor can create plasmas heated to 70 million degrees C. For its commercial device, the company says it needs to heat plasmas to 1 billion degrees C.
Binderbauer told Axios that he is aiming to raise another $50 million before the round closes later this summer. The company is hoping to put electrons on the grid sometime in the early 2030s.
