🚀Just goes on to show— The best way to invest in Hunanoid is Neocloud and HBM.
Physical AI will also require same Compute, Memory, Power and Edge inferencing!!
$NBIS$HUT8$CIFR$IREN$DRAM$Mu$SNDK$CCXI
Case— Figure is partnering with Nscale to deploy up to 100,000
Clearly shows AI demand and Cash to fund Capex exists.. it’s just execution hurdles in supply chain to scale that curbs growth rates!
If in doubt , just look at $NVDA FCF at $127 billion and has $100 billion cash at hand.
$NVDA will and can backstop large deals to keep the AI
🚀 $AVGO - Q3 AI semiconductor revenue of $16.7B, up 221% YoY and 54% QoQ
Why the ASIC story is the whole story now for $AVGO ??
🚨AI semis are now 56% of total revenue. 2 years ago Broadcom was a networking and-software company with an AI pivot. That framing is changed upside. h
🚨Elon Musk at G20 is betting on ‘Humanoids’ as the next big-theme—
“There will be at least a billion robots in 10 years, each producing at least 5x the output of a human”
Morgan Stanley says robots are a $5 TRILLION market by 2050.
Here are 10 stocks/ ETFs that ‘ll
🚨This is a ‘real risk’ for small/upcoming with Hyperscalers CDS rising, higher yields as new normal and data center being the ‘political flashpoint’
Am bullish on Neoclouds who has the cash on books, backstop from $NVDA and order visibility line $NBIS
NFA
🚀Strong NUMBERS by $CRDO :
∙Revenue $479M, up 10% sequentially and 115% YoY — seventh straight quarter of triple-digit growth. EPS $1.20
∙Non-GAAP gross margin 68%, operating margin 48.2%, net income a record $236.3M
📈Optics as % of FY27 revenue—
Management reiterated h
🚨 The most overlooked ‘AI trade’ just got its own ETF !! Goldman calls it “the next memory.” 🧵
$CAPA — the first US-listed capacitor & MLCC ETF — went live Aug 26
Why?
A single Nvidia GB300 AI server needs 30,000 MLCCs. A full rack? 320,000. The next-gen Rubin rack? 570,000 h
🚨This is extremely bullish for $NBIS given the shift open weights like https://t.co/NCZdrVZBYY blockbuster release last week !!
AI World is rapidly shifting to Openweight models. On Vercel’s production gateway, open-weight models handled 29% of tokens but under 4% of spend.
