The sector rebound has not provided a clear reason to abandon my thesis: delivering its backlog requires gigawatts of additional capacity and, by my estimate, more than $100 billion of buildout.
That expansion means substantial funding needs, with debt and potential shareholder dilution to absorb. My concern remains how CoreWeave finances those commitments, and whether the likely fallback is even greater reliance on NVIDIA.
A rebound alone does not show that these pressures have eased. Until there is clearer evidence that the underlying funding problem has changed, I don’t see a convincing reason to call the thesis broken.