What to know about Nvidia's massive spending spree

Watch on YouTube ↗  |  January 02, 2026 at 19:43  |  4:22  |  CNBC
Speakers
Kristina Partsinevelos — Markets Reporter, CNBC

Summary

CNBC's Kristina Partsinevelos explains Nvidia's $30 billion inference-focused spending spree, including deals for Groq, Fabricas, and a reported AI21 Labs acquisition. The strategy aims to defend Nvidia's AI dominance as inference fragments and opens the door to competitors. She notes Nvidia stock has been flat for about six months due to profit-taking and a lack of next catalyst, with investors rotating into other AI picks and shovels like memory.

  • Nvidia is spending up to $23 billion in under four months on inference.
  • Inference is the AI model deployment phase after training.
  • Nvidia owns about 90% of AI training but inference is fragmenting.
  • Deals include Groq, Fabricas, and reportedly AI21 Labs.
  • Nvidia has strong free cash flow and large buybacks.
  • Nvidia stock has been flat for about six months.
  • Investors are rotating into memory and other AI picks and shovels.
  • Competitors like Google, Amazon, and custom-chip makers may benefit.
Ideas
Kristina Partsinevelos Markets Reporter, CNBC 1:02
Nvidia spends big to defend inference dominance.
Nvidia is making a massive, potentially $23 billion spending push in under four months—via deals for Groq, Fabricas, and reportedly AI21 Labs—to dominate inference, the next phase of AI, and lock down talent and technology before competitors. It owns about 90% of AI training, but inference is fragmenting and cheaper to run, opening the door to Google, Amazon, and custom-chip rivals. The stock has been flat for about six months as investors take profits in AI plays and await the next catalyst, though Nvidia has roughly $80 billion in free cash flow, estimated to reach $96 billion, and has spent about $52 billion on buybacks.
NVDA HIGH
Kristina Partsinevelos Markets Reporter, CNBC 3:25
AI investors rotate into memory.
Kristina says that with Nvidia lacking a next big catalyst and investors taking profits in many AI plays, stockholders are rotating into other AI picks and shovels, especially memory, which has already seen a gigantic surge.
Memory MED
Kristina Partsinevelos Markets Reporter, CNBC 3:32
Investors rotate into AI picks and shovels.
Kristina says the upside for many stockholders is to move into other picks and shovels of the AI trade because Nvidia lacks a near-term catalyst and there is profit-taking in the large AI plays.
AI picks and shovels MED
Up Next
1:41
Mamdani Rides Train to Work a Day After His Inauguration
Bloomberg Markets · Jan 02
4:01
Why Ed Yardeni is still bullish on stocks overall in 2026
CNBC · Jan 02
4:50
Anthropic's strategy to outmaneuver big-spending rival OpenAI
CNBC · Jan 02
3:23
Why G Squared's Greene sees Verizon as a safe harbor stock in 2026
CNBC · Jan 02
4:42
Anticipating Tesla to scale-up CyberCab production in 2026, says Canaccord Genuity's Gianarikas
CNBC · Jan 02

This CNBC video, published January 02, 2026, features Kristina Partsinevelos discussing NVDA, Memory, AI picks and shovels. 3 trade ideas extracted by AI with direction and confidence scoring.

Speakers: Kristina Partsinevelos  · Tickers: NVDA, Memory, AI picks and shovels