@meloncurls21 Compute being sold out that far out makes you wonder if we're building too fast without thinking about the real-world limits. Power and land aren't infinite.
AI is not only getting bigger.
It is also getting cheaper.
TurboVec, built on Google Research’s TurboQuant, can shrink AI vector memory from 31GB to 4GB.
That means:
Less memory.
Faster search.
Offline AI.
No expensive GPU cluster.
No cloud dependency.
Everyone is watching bigger models.
But the real shift may be cost reduction.
If AI becomes cheaper to run, adoption gets faster.
The stocks to watch:
$GOOG $NVDA $MSFT $AMZN $META $AMD $AVGO
My view:
The next AI trade is not only about who builds the biggest model.
It is about who makes AI cheaper, faster and easier to deploy.
So SpaceX and Google are doing this deal, and it’s way bigger than another AI headline. 🚀
Google is paying SpaceX $920M per month for AI compute until 2029.
That includes about 110,000 NVIDIA GPUs in SpaceX data centers.
This really shows me one thing:
Even Google doesn’t have enough compute.
The real fight in AI isn’t just chatbots or models anymore.
It’s about who controls:
GPUs
Data centers
Cloud capacity
Power
Infrastructure
SpaceX was always rockets and satellites, but now they’re turning AI compute into a cash machine.
After the Anthropic deal, SpaceX’s AI compute contracts could make around $26B a year.
That’s wild.
Stocks I’m keeping an eye on:
$GOOG
$NVDA
$AMD
$AVGO
$MSFT
$AMZN
$ORCL
$TSLA
$TSLA isn’t direct SpaceX, but the whole Elon AI infrastructure story will still matter.
My take is simple:
AI isn’t just software anymore.
It’s an infrastructure arms race.
Don’t chase hype.
Follow the compute.
$MU is quickly turning into a major player in the AI infrastructure space.
The basic idea is pretty straightforward:
AI models keep getting bigger.
Data centers are growing fast.
Memory bandwidth is hitting a wall.
That's where Micron comes in.
Micron deals with HBM, DRAM, NAND, and storage for data centers—all essential for AI servers and cloud setups.
That's why I don't see $MU as just a regular memory cycle stock anymore.
To me, it's an AI infrastructure supplier now.
Stocks I'm keeping an eye on:
$MU — AI memory
$NVDA — AI processing
$AVGO — networking / custom silicon
$TSM — chip manufacturing
$ASML — lithography gear that's a bottleneck
My plan:
Don't blindly chase after strong moves.
Wait for dips.
See if buyers show up at key support levels.
Only jump in when the risk-reward feels right.
Not advice.
🚨 $AVGO is taking a hit, but the AI infrastructure story is still solid.
Price now: $385.75
Today: -7.92%
The market is just reacting to expectations.
But the core business logic hasn't changed:
AI data centers need more than just $NVDA GPUs.
They need custom chips, networking, switches, optical gear, and infrastructure software.
That's where Broadcom comes in.
$NVDA builds the AI engine.
$AVGO connects and scales the whole AI factory.
So I don't see a broken company here.
I see a stock where I'm waiting for a better entry.
My watchlist:
$AVGO / $NVDA / $TSM / $ASML / $MU
Don't jump in.
Wait for support.
Let fear create the opportunity. 🚀
Not financial advice.
@meloncurls21 100x by 2030 feels like a headline grabber more than a grounded forecast. AI revenue is possible but that kind of growth assumes a lot of stars align.