AI “Bottleneck Trade” Nearing End, Early SpaceX Investor Says Market Focus Will Shift to Long-Term Value

BlockBeats News, June 17th, the hot trade surrounding the artificial intelligence supply chain shortage may be coming to an end, as some investors are starting to shift from seeking the next bottleneck to filtering out companies that still have long-term competitiveness after the AI infrastructure cycle.

Altreides Management Managing Partner and early SpaceX investor Gavin Baker said in an interview with TBPN that over the past year, the market has been enthusiastic about chasing bottleneck assets in AI development, including DRAM, storage chips, and key material suppliers. However, he believes that such "AI bottleneck trades" are nearing an end.

Baker pointed out that Japan's Ajinomoto has refused to raise the price of a key chip packaging insulating material, indicating that the pricing power of some supply chain bottlenecks may have begun to ease. This material is used to connect the processor and chip in the packaging layer, which had previously attracted investor attention due to the surge in demand for AI chips.

He said that the previous market game was to find the "next bottleneck," but the more important question in the next stage is which companies can still maintain enduring franchise value after these bottlenecks dissipate.

This view has cooled the recent surge in storage and AI material stocks. Storage stocks such as Micron and SanDisk have risen sharply this year, driven by AI data center capital expenditures, HBM demand, and a reevaluation of supply and demand brought about by long-term procurement agreements. However, as stock prices have risen rapidly, the market has begun to discuss whether the trades are becoming overcrowded.

Baker also mentioned that the next focus of AI infrastructure may shift to the practical implementation of computing power and electricity. He referred to SpaceX's potential in ground and orbital AI data centers, stating that the market will focus on its speed in increasing "gigawatts" of computing power supply.

This means that AI trades may shift from simply betting on shortages to a broader infrastructure competition: whoever can quickly acquire land, power, GPUs, and data center capacity may be in a more favorable position in the next stage of the AI investment cycle.