The November 2022 launch of OpenAI’s ChatGPT sparked a global AI race, reaching one million users in just five days. This milestone set off a frenzy among tech giants like Microsoft, Alphabet, Amazon, and Meta, which ramped up capital expenditures to build the infrastructure needed to power AI’s transformative potential. Data centers, GPUs, and custom chips now underpin what many consider the next ‘industrial revolution’.
Advancements in AI models have accelerated at an astonishing pace, fueled by exponential growth in computational infrastructure. From GPT-2’s 1.5 billion parameters to GPT-4’s 1.8 trillion, training these models demands unprecedented levels of computing power and energy. Nvidia, the GPU leader, is at the heart of this revolution. GPT-4’s training alone required an estimated 25,000 Nvidia A100 GPUs running simultaneously, contributing to Nvidia’s meteoric rise as the most valuable company in the world. Yet even Nvidia can’t meet surging demand; tech giants like Amazon and Microsoft are scrambling to secure GPUs to support their AI ambitions.
Capital expenditures reflect the urgency: Microsoft’s quarterly capex now surpasses its pre-2020 annual spending, driven by Azure AI demand. Alphabet’s Q3 2024 capex hit $13 billion, a +62% YoY increase, while Amazon reported an +81% surge in spending. Meanwhile, LLM developers like OpenAI and Anthropic have raised billions to stay ahead, with OpenAI projecting $5 billion in ARR by the end of 2024.
Private markets are also reaping rewards. Cloud infrastructure leaders like CoreWeave and Lambda as well as data platforms like VAST Data and Databricks are thriving on AI-driven growth. G Squared’s strategic investments in these sectors position our portfolios to capitalize on the AI revolution, fueling innovation while generating value across partnerships. The race is on.
