You don't "just hook it up to the grid". You have to work with the local utility company or companies to negotiate pricing, SLAs, and safety concerns (eg: shutting off power when linemen are working). Helion's fusion reactors are small enough to fit in a couple shipping containers. The idea is that the units can be mass produced in a factory, then transported to data centers and replace the grid hookup as the primary power source. Another potential market is stuff like remote mining operations which currently use diesel generators. The cost of transporting fuel to such sites can be quite high. A fusion reactor, while more expensive up-front, could be cheaper in the long run. And unlike solar cells, the reactor can be easily moved offsite when the operation is complete.
But you do have to compete with grid hookups for pricing in data centers.
I definitely buy that there's a market for small power plants that cost too much to be competitive on the grid (in some sense that's literally what generators already are), but I don't see how datacenters are in that market.
> but I don't see how datacenters are in that market
Are you kidding? If datacenters aren't the majority of that market, they will be soon, and their share will keep growing on the short term. Only factories match their consumption, every other wholesale energy buyer is relatively tiny.
But that market will pay at most a 3x to 5x markup on the grid price. There exists a smaller one formed by far from the grid activities that will pay a much higher price.
Wholesale buyers are willing to pay 3x to 5x (depends on the location) more than electricity distributors.
Electricity distributors may add a markup of over 10x on resale. For many reasons, it's easier to get energy out of them so you just won't pay the full ~10x markup on the wholesale market.
> But you do have to compete with grid hookups for pricing in data centers.
You do, unless you sign a contract with a large datacenter operator that's willing to pay above market price for the PR benefit/green energy credits of a working fusion reactor. Which presumably is the point of this.
And realize this also depends on how much the competition costs. If renewables and storage continue to get cheaper, fusion that "works" in an economic sense could transition back to "doesn't work".
It's the fate of most technologies to fail this way, so it shouldn't be surprising (or a source of dishonor) if the same thing happens to fusion. That doesn't mean money spent developing fusion is necessarily wasted; the size of the energy market -- a quadrillion dollars or so over the 21st century -- justifies longshot investments.
Bear in mind that Helion's reactor is 50MW, about the size of three large wind turbines. If it works they'll be mass-producing them in factories and shipping them by rail, and they'll have a declining cost curve just like anything else.