@AshCrypto spacex valued like an entire asset class. tells you what markets do with clear narratives
crypto still priced like a trade, not an infrastructure layer
when data becomes an income stream not a byproduct, that is where $DITAU starts to get interesting
@cryptofergani the loud part is btc
the quiet part is data
institutions don’t just buy coins, they buy information flow
watch who corners ai + onchain data rails next cycle, that is where $DITAU fits in
Everyone chased compute multiple expansion and missed the capex shift to storage and memory. Next leg is who actually deploys and operates all this capacity, not who fabbed it.
🚨 IF YOU INVESTED $100K INTO THESE AI CHIP STOCKS 15 MONTHS AGO YOU WOULD HAVE OVER $1.5 MILLION TODAY.
Here is what drove that return, stock by stock.
SanDisk went from a $4 billion company to $321 billion.
Western Digital went from $10 billion to $269 billion.
Micron crossed $1 trillion, now sitting at $1.25 trillion.
Seagate moved from $14 billion to $250 billion.
Intel climbed from $89 billion to $652 billion.
AI data centers are expected to consume 70% of the entire global memory chip supply this year.
Micron has confirmed it is sold out of memory supply for all of 2026.
Western Digital's high capacity hard drives are fully committed to hyperscaler customers through the rest of the fiscal year. SanDisk locked in its NAND supply by extending its manufacturing deal with Kioxia all the way to 2034.
For most of the past decade, Nvidia was the only chip stock that mattered, while memory and storage names traded sideways.
That has completely flipped. The AI trade did not slow down. It moved from compute straight into memory.
How much profit the Mag 7 are expected to generate over the next 3 years:
• $NVDA ~$1.1T
• $GOOGL ~$621B
• $MSFT ~$543B
• $AAPL ~$507B
• $AMZN ~$406B
• $META ~$318B
• $TSLA ~$25B
@cryptorover funny part is most of that flows to the same narrow set of tickers again and again. real asymmetry is usually off-index, before the crowd notices it exists
@TrendSpider funny part is nvda is the billboard, not the buildout
the quiet stuff is in the infra names like $GPUS where the actual racks, bots, and power scaling live
@GrindeOptions mu at ath is just the market finally pricing in that memory is ai infra, not a cyclical afterthought. if you like that thesis, the edge might be in smaller ai infra plays like $GPUS where that same repricing hasn’t happened yet
@fammetaX funny part is the real asymmetry isn’t even in the goog names. it’s in the little infra names that feed them. people will notice $GPUS way later, like always
@KobeissiLetter energy volatility now basically trades off drones and headlines. if crude keeps fading, margin pressure on high-power data centers eases a bit, but ai buildout is still bottlenecked on power delivery and grid upgrades more than feedstock prices
@WatcherGuru If private markets are willing to put $2T on SpaceX, public markets still feel cheap on anything tied to physical AI buildout. Space, data centers, robotics, energy. Market still acting like it is only about $NVDA multiples.
@cryptorover that was the first real reminder in a while that geopolitics can still gap-risk equities. everyone’s long long-vol but short actual event risk. if this keeps repeating, physical stuff like energy, defense, and AI infra get bid over pure software multiples.
@JasonL_Capital that book-to-bill >4x on datacenter is wild. market is still pricing AI optics like a cyclical, not like core infra. feels similar to how people treated GPUs pre-2023. watching how this spills into the broader AI hardware stack like $GPUS-type names too
@TaraBull Rates + AI tightening at the same time.
Positioning was max long, leverage high, everyone crowded in the same trade. When bonds, gold, equities and $BTC all puke together, it’s not narrative. It’s de-risking and funding.
This is what forced sellers look like.
@TheTechInvest nice list. everyone crowds the GPU and hyperscaler layer, but the real squeeze is coming in physical buildout. i’m watching $GPUS alongside $NVDA for who actually scales racks, power and robotics in these boxes, not just sells chips
@StockSavvyShay cool off was overdue. ai trade got crowded short term. if data center capex or yields keep biting, next leg probably rotates to physical infra plays like $GPUS that don’t need perfection priced in
@StockSavvyShay agree on the custom silicon shift, but i’d look at who scales the physical side too. when model caps hit trillions, data center buildout and ai-focused lessors like $GPUS can see a delayed but violent catch-up bid as that capex actually gets deployed.