Compute
The computational resources (processing power, memory, time) required to train or run AI models. Compute is measured in FLOPs (floating-point operations) and is a primary constraint and cost in AI development.
Why It Matters
Compute is the fundamental currency of AI progress. Access to compute determines who can build frontier models and at what cost.
Example
Training GPT-4 is estimated to have required ~$100 million in compute costs, using thousands of GPUs running for months.
Think of it like...
Like the fuel needed for a rocket launch — bigger rockets (models) need exponentially more fuel (compute), and the cost is a major constraint on what you can build.
Related Terms
GPU
Graphics Processing Unit — originally designed for rendering graphics, GPUs excel at the parallel mathematical operations needed for training and running AI models. They are the primary hardware for modern AI.
TPU
Tensor Processing Unit — Google's custom-designed chip specifically optimized for machine learning workloads. TPUs are designed for matrix operations that are fundamental to neural network computation.
FLOPS
Floating Point Operations Per Second — a measure of computing speed that quantifies how many mathematical calculations a processor can perform each second. Used to measure AI hardware performance.
Cloud Computing
On-demand access to computing resources (servers, storage, databases, AI services) over the internet. Cloud providers like AWS, Azure, and GCP offer scalable infrastructure without owning physical hardware.
Scaling Laws
Empirical findings showing predictable relationships between model performance and factors like model size (parameters), dataset size, and compute budget. Performance improves as a power law with these factors.