MoneyThe story, in brief
Nvidia says Groq racks will be online this year following $20 billion purchase
$20B. Nvidia just acquired Groq—and says inference racks hit production this year.
The KeyNews take
Why it matters
A major acquisition signals Nvidia's strategic bet on low-latency inference as a separate compute category, reshaping the inference market and raising questions about vertical integration in AI hardware.
The key facts
4 to knowNvidia acquires Groq for $20 billion
Groq racks expected to be online this year (2026)
Strategic focus: low-latency inference capabilities
Groq's inference-optimized architecture vs. Nvidia's general-purpose GPUs
Go to the source
CNBC Technologycnbc.com
Publisher excerpt: Nvidia's race to manufacture Groq chips and make them available to customers highlights the growing importance in AI of low-latency inference.