Google restricts Meta from using Gemini, citing insufficient computing power. This news points to the physical bottleneck of the AI arms race — the computing power supply of hyperscale cloud providers has fallen below market demand, and they cannot even fill in for each other.

For the crypto market, this means two things: First, the logic of AI data center capital expenditure continues to strengthen, and related tokenized assets (such as RWA, computing power tokens) may gain narrative premium; Second, the computing power squeeze will push up GPU leasing costs, indirectly compressing the profit margins of miners and AI projects.

But the risk is that the market may over-extrapolate linearly — capital expenditure does not equal immediate output, data center construction cycles are long, cash flow returns are uncertain, and crypto-native computing power projects generally lack actual revenue support.

When AI giants start to "lock doors" on each other, computing power shifts from infinite supply to scarce resources. The crypto world needs to reassess which projects truly have computing power moats and which are just riding the narrative.

$ai #rwa #ai #区块链 #crypto market