GPU

Definitions

1. Graphics Processing Unit — a processor originally designed for rendering graphics but whose massively parallel architecture turned out to be ideal for machine learning workloads. NVIDIA's GPUs became the picks and shovels of the AI gold rush, making it one of the most valuable companies on Earth.

   In plain English: A computer chip originally designed for video games that turned out to be perfect for AI — now the most sought-after hardware in the tech industry.

2. The AI GPU shortage (2023-2025) made NVIDIA's H100 chips the most sought-after commodity in tech, with wait times exceeding 6 months. This scarcity drove the GPU-as-a-service cloud market and motivated companies like Google (TPU) and Amazon (Trainium) to build custom AI accelerators.

   Example: 'We're on a 4-month waitlist for H100 GPUs. In the meantime, we're training on rented A100s at $3/hour and optimizing our model to need fewer GPUs.'

   Source: scarcity / market dynamics

Etymology

1999

NVIDIA coins 'GPU' (Graphics Processing Unit) with the GeForce 256, marketing it as the world's first GPU

2007

NVIDIA releases CUDA, enabling general-purpose computing on GPUs and inadvertently creating the hardware foundation for deep learning

2023

The AI boom makes NVIDIA's GPUs the most sought-after chips in the world; the company's market cap exceeds $1 trillion

Related Terms

• Inference
• Quantization
• Overclocking
• CUDA Core
• FLOPS
• GPU Poor
• Groq Chip
• Jensen Huang
• Neural Processor