Despite a historic $600 billion market cap loss after DeepSeek’s AI breakthrough, NVIDIA (NVDA) has received strong support from Meta and Microsoft, signaling a bullish future for its AI chip business.
China’s DeepSeek claimed it trained an AI model matching GPT-4 for under $6 million, compared to OpenAI’s $100M+ training cost.
This sparked fears that US companies might reduce spending on NVIDIA’s GPUs, cutting into its AI revenue stream.
Meta CFO Susan Li announced a 66% increase in AI capital expenditures in 2025, reaching $65 billion.
CEO Mark Zuckerberg emphasized that AI efficiency gains don’t reduce GPU demand—instead, they increase AI workloads, requiring more computing power.
Microsoft CEO Satya Nadella sees AI cost reductions driving exponential GPU demand, citing Jevons Paradox—as AI gets cheaper, companies will use even more GPUs.
AI infrastructure spending among top hyperscalers (Amazon, Alphabet, Meta, Microsoft) is now projected to hit $317 billion in 2025, up from previous estimates.
Despite concerns, NVIDIA’s median target price remains at $175—a 45% upside from its current price. With AI demand surging, the long-term outlook remains strong.