NVIDIA has already written the answer on its face—it's just that the market prefers to believe it only after the rally is over
On August 16, Serenity pointed out that AI investment is actually not as complicated as many people think, because NVIDIA and Jensen Huang have been steadily sending signals all along.
The real issue is this:
The market isn’t unable to see opportunities—it just isn’t willing to believe them.
Only when the stock price starts to surge do people suddenly realize—so someone had already laid the groundwork early.
Looking back at the AI hardware wave in 2025:
When NVIDIA moved ahead with core optical-interconnect production capacity such as EML and laser emitters, the market didn’t buy it at the time.
Many people believed:
“Photonics is hype.”
“Just like quantum computing—it’s only a capital story.”
Even related companies were at one point labeled as “a bubble” or “a scam.”
But a year later, the market cast a new vote with prices:
LITE rose 678%, AAOI climbed 475%, COHR gained 261%, and AXTI surged by as much as 3,844%.
What had once been mocked eventually became a new hotspot for AI infrastructure.
And now, a similar storyline is playing out again.
NVIDIA is positioning early around the next generation of AI infrastructure:
locking in CW/EML capacity, pushing the 800V power architecture, accelerating the rollout of CPO (co-packaged optics) technology, and elevating “physical AI” and robotics to a new strategic height.
But the market’s current voices still sound familiar:
“CW companies are just Meme stocks.”
“800V is still too early.”
“CPO commercialization is still far off.”
“Humanoid robots can’t make money.”
History keeps proving this:
The truly big opportunities rarely appear when everyone agrees.
More often than not, industry leaders have already told you the answer—it’s just that the market needs to wait until results are realized and the stock price rises before it’s willing to admit it missed out.
The core of the next phase of competition in AI isn’t only the models, but the underlying infrastructure—computing power, electricity, optical interconnects, robotics, and more.
Smart money doesn’t chase after a hype cycle forms; it understands the next trend when everyone else is questioning it.
On August 16, Serenity pointed out that AI investment is actually not as complicated as many people think, because NVIDIA and Jensen Huang have been steadily sending signals all along.
The real issue is this:
The market isn’t unable to see opportunities—it just isn’t willing to believe them.
Only when the stock price starts to surge do people suddenly realize—so someone had already laid the groundwork early.
Looking back at the AI hardware wave in 2025:
When NVIDIA moved ahead with core optical-interconnect production capacity such as EML and laser emitters, the market didn’t buy it at the time.
Many people believed:
“Photonics is hype.”
“Just like quantum computing—it’s only a capital story.”
Even related companies were at one point labeled as “a bubble” or “a scam.”
But a year later, the market cast a new vote with prices:
LITE rose 678%, AAOI climbed 475%, COHR gained 261%, and AXTI surged by as much as 3,844%.
What had once been mocked eventually became a new hotspot for AI infrastructure.
And now, a similar storyline is playing out again.
NVIDIA is positioning early around the next generation of AI infrastructure:
locking in CW/EML capacity, pushing the 800V power architecture, accelerating the rollout of CPO (co-packaged optics) technology, and elevating “physical AI” and robotics to a new strategic height.
But the market’s current voices still sound familiar:
“CW companies are just Meme stocks.”
“800V is still too early.”
“CPO commercialization is still far off.”
“Humanoid robots can’t make money.”
History keeps proving this:
The truly big opportunities rarely appear when everyone agrees.
More often than not, industry leaders have already told you the answer—it’s just that the market needs to wait until results are realized and the stock price rises before it’s willing to admit it missed out.
The core of the next phase of competition in AI isn’t only the models, but the underlying infrastructure—computing power, electricity, optical interconnects, robotics, and more.
Smart money doesn’t chase after a hype cycle forms; it understands the next trend when everyone else is questioning it.
