TickerTain
TickerTain
NewsroomShortsPortfolioConvergence
NewsroomShortsPortfolioConvergence
$ANTHROPIC·$MA····$INTC····$BLUE-ORIGIN·$SPCX····$CRWV····$CRM····$MSFT····$NVDA····$ORCL····$CURSOR·$AAPL····$OPENAI·$AMZN····$UBER····$GOOGL····$META····$TSLA····$DATABRICKS·$PERPLEXITY·$LYFT····$NBIS····$TSM····$LITE····$ANDURIL·
$ANTHROPIC·$MA····$INTC····$BLUE-ORIGIN·$SPCX····$CRWV····$CRM····$MSFT····$NVDA····$ORCL····$CURSOR·$AAPL····$OPENAI·$AMZN····$UBER····$GOOGL····$META····$TSLA····$DATABRICKS·$PERPLEXITY·$LYFT····$NBIS····$TSM····$LITE····$ANDURIL·
←

sha nandy

T3 · host / generalist

Leads AWS technology strategy for AI/ML workloads; predicts 90% inference cost reduction via purpose-built chips and specialized small models.

1 call·1 name·100% bull·last heard 10 months ago·The Information
track record

no scored calls yet — needs a stated position or a categorical verdict, with a matured window vs SPY

top calls

highest conviction · one per company
1sthigh conviction
$AMZNAmazon

AWS director Sha Nandy predicts 90% inference cost drop within 1-2 years

Inference costs will drop 90% due to shift from training to inference, diverse purpose-built chips, and smaller specialized models like Nova micro unlocking agentic workflows.

The Information2025-11episode →

most discussed · click a bar to filter

  • $AMZN

recurring themes

  • AI Infrastructure1
1 total
$AMZN
···
Amazon
HIGHsha nandy·The Information·10 months ago·General Catalyst's Novel VC Fund, Creator Economy Shift, AI Inference Cost Prediction | Nov 19, 2025
AWS director Sha Nandy predicts 90% inference cost drop within 1-2 years
Inference costs will drop 90% due to shift from training to inference, diverse purpose-built chips, and smaller specialized models like Nova micro unlocking agentic workflows.
"We see inference costs dropping by 90%. Over the next year or two... the consumption is shifting to usage and usage is what we largely call inference... you're also seeing many di…"
21:13
8
AI Infrastructuretailwind
Inference costs to drop 90% in 1-2 years as workloads shift to specialized chips and small models
Transition from training to inference, purpose-built silicon (Trainium/Inferentia), and agentic routing to tiny models like Nova micro will radically lower cost per token, unlocking new use cases.