- In The Loop AI
- Posts
- NVIDIA exceeds quarterly expectations
In Today’s Issue:
NVIDIA exceeds quarterly expectations
DeepSeek Updates R1 AI Model
Read time: 2.31412 minutes
NVIDIA exceeds quarterly expectations
NVIDIA reported $44.06 billion in fiscal Q1 revenue, up 69% year-over-year, driven largely by demand for its AI chips, despite a $2.5 billion hit from U.S. export restrictions to China. The company’s data center division, which powers most AI workloads, brought in $39.1 billion, up 73% from last year. Net income reached $19.89 billion, even after a $4.5 billion inventory charge on China-bound chips rendered unsellable by new rules from the Trump administration. NVIDIA projects $45 billion in Q2 revenue, including an $8 billion loss from the H20 chip ban. Meanwhile, demand for its next-gen Blackwell chips, key to powering large-scale AI models, is expected to drive nearly $100 billion in sales this fiscal year alone. While its China outlook remains uncertain, Nvidia continues to benefit from surging U.S. AI investment. Microsoft, Google, Meta, and Amazon are expected to spend over $345 billion on AI infrastructure in 2025.
Deepseek updates R1 AI model
Chinese startup DeepSeek has released an upgraded version of its R1 reasoning model on Hugging Face under an MIT license, enabling commercial use. The update, described as a “minor trial upgrade,” was announced Wednesday via WeChat, though no technical details were shared. The original R1, launched in January, drew global attention for outperforming Western models at a fraction of the cost, sparking a selloff in AI stocks. The R1 update maintains its massive 685 billion parameter size, making it one of the largest publicly released models to date, though it remains impractical for most consumer hardware.
How was today's newsletter? |
Reply