Speakerguest

Dylan Patel

@dylan522p / appears on iltb, iltb, tweets, iltb, dwarkesh, tweets, dwarkesh, dwarkesh, tweets, iltb, iltb, tweets, iltb, iltb, iltb, dwarkesh, iltb, tweets, dwarkesh, iltb, semianalysis, dwarkesh, tweets, iltb, tweets, iltb, tweets, iltb, semianalysis, tweets, iltb, iltb, dwarkesh, iltb, tweets, semianalysis, tweets, dwarkesh, iltb, tweets

calls12traced speaker calls

avg P&L-0.3%per call

win rate75%resolved calls

total P&L-3.5%expert engine

Verified record

39credibility

win rate25%1W · 3L

resolved4settled on TRACE

scored calls19published & scoreable

Speaker record

episodelongNVDAtweets

NY's data center moratorium is symbolic noise; AI compute demand is too large and global for any state to stop

I would like to see them TRY STOP THIS

+2.21%open horizon

episodelongNVDAtweets

SemiAnalysis Tokenomics Model projects surging AI compute demand, and Nvidia supplies the GPUs that make frontier model training possible.

here are the AI compute projections from our new Tokenomics Model

+1.23%open horizon

episodelongMETAtweets

Meta is the only hyperscaler world-class at data, talent, AND compute — uniquely positioned to close the AI frontier gap with OpenAI/Anthropic.

Meta is the only hyperscaler/neolab on track to be world class at all three and therefore has the best chance at catching up with Anthropic/OpenAI

+1.22%open horizon

[00:00]longMAGSsemianalysis

Agent-readable speaker

{
  "type": "trace.speaker",
  "handle": "dylan522p",
  "name": "Dylan Patel",
   [
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    ,
    
  ],
   {
     
  }
}

AISO Market Calls classifies public statements from public sources. A classification such as long thesis, short thesis, or explicit call describes the content of the public statement and does not prove that the speaker actually entered, exited, or held a position. Market performance shown is calculated from public market data around the source publication or timestamp. It is not investment advice.

Dylan Patel

NY's data center moratorium is symbolic noise; AI compute demand is too large and global for any state to stop

SemiAnalysis Tokenomics Model projects surging AI compute demand, and Nvidia supplies the GPUs that make frontier model training possible.

Meta is the only hyperscaler world-class at data, talent, AND compute — uniquely positioned to close the AI frontier gap with OpenAI/Anthropic.

Mag 7 already worth 1.8x Europe; AI value will compound the same way computers did

Captured AI value accrues to firms with market power who hold price as costs collapse

AI output is invisible in macro data while costs are visible, so the AI-heavy index gets misread cheap.

NVIDIA margins stay fat as GPU prices rise and useful lives extend far beyond 5 years.

ASML completely sold out; Carl Zeiss is the bottleneck on expanding lithography supply.

OpenAI is far from cooked; aggressive compute buying lets it serve the demand Anthropic can't.

Memory prices double or triple again from here; capacity can't grow fast enough until 2028.

TSMC could spend $100B on CapEx by 2028, whipping its equipment suppliers harder than anyone expects.

Semiconductor equipment makers get the tail-whip from TSMC's surging CapEx; still underappreciated.

Anthropic margins float at a 72%+ floor as token demand vastly outstrips compute.

CPUs completely sold out as RL training environments and AI-generated app sprawl drive demand.

Robotics gets few-shot learning breakthroughs in 6-18 months as cheap implementation unlocks new models.

IREN underpriced as scarcity-driven rental pricing power lifts all Neocloud economics

DRAM and NAND memory prices going parabolic on AI demand, a structural supply squeeze

Anthropic demand exploding, ARR tripled in a quarter as agentic coding consumes software dev

Nebius mispriced as a Neocloud beneficiary of rental tightness while sentiment stays anchored to oversupply

Neocloud stocks are mispriced, sitting at lows while rental pricing power surges and useful life extends

Memory prices keep doubling as KV cache and HBM demand swamp slow-to-build fabs. Long memory.

ASML is the ultimate AI bottleneck by 2028-29; only it makes EUV and can't scale fast. Long ASML.

TSMC's tiny 3nm CapEx slice converts into ~$160B/yr of NVIDIA earnings; foundry holds the keys. Long TSMC.

Apple gets squeezed out of leading nodes and hit hard by memory price spikes; phone volumes fall.

Memory makers SK Hynix and Samsung capture the margin as AI memory triples and they sign long-term deals.

Modular DC assembly moves value to Taiwan factory integrators; Foxconn and Delta capture more of each buildout.

AI model vendor margins go up this year because they're so capacity-constrained they must destroy demand.

Long AI timelines give China years to scale fabs, and Alibaba is the tradable winner of domestic compute.

Electrician scarcity gates the AI buildout, and the MEP contractor that commands that labor prices the premium into backlog.

Google woke up late and is now absurdly AGI-pilled, buying power, land and compute aggressively.

Reach & recent activity