UEX is not a competition over the number of assets; it is a competition over user value, asset efficiency, and trading experience.

My AI investments haven't paid off yet, so are tech giants really sitting on excess computing power?

Is it still the golden age?

Good technology deserves commercial success.

如果微软、谷歌、亚马逊决定停止每季度300亿美元的GPU采购，整条供应链都将崩塌。

For investors focused on chip stocks, the current situation can be summed up in one sentence: uncertainty is extremely high, and short-term volatility will intensify.

"Of all assets in existence today, BTC offers one of the widest margins of safety, even if the ultimate low is roughly 15-20% below current prices."

The true direction still needs to be validated by the earnings season.

Meta's plan to sell excess computing power shattered the belief in "absolute scarcity of computing power," sending its stock price soaring 10% in a single day, its best performance of the year, while AI hardware stocks like chipmakers suffered heavy losses.

Gemini and DeepSeek write underdog storylines, while Grok and Qwen nail hot-favorite matches with small scorelines; ChatGPT and Claude are better suited for match process analysis.

Iran is preparing to sell oil to all countries except Israel.

It wasn’t luck; it turned “gambling” into a robust system.

Murata Remains Core Long Position, but Hedge Funds Begin Seeking Highly Valued Short Targets

As AI Agents become a productive force that can be invoked, traded, and combined, OKX chooses to continuously build the economic infrastructure for it.

There was a prevailing narrative in the AI space: that improved inference efficiency would kill the demand for chips. However, the reality is the opposite—the cheaper AI becomes, the more valuable physical compute power gets. Every price cut that model developers compete on ultimately flows into the pockets of chipmakers and foundries.

AI demand is driving up the popularity of PCBs, optical interconnects, and SiC