What it is
TOPS, or Trillions of Operations Per Second, measures the rate at which a processor can perform integer operations, often used for inference tasks in AI where lower precision calculations are sufficient. FLOPS, or Floating-Point Operations Per Second, quantifies the speed of floating-point computations, which are critical for scientific simulations and AI model training requiring high precision. Both metrics indicate a chip's raw processing power.
Chip manufacturers often advertise their products' performance using TOPS for inference chips and FLOPS for training chips, particularly GPUs and AI accelerators. These metrics help buyers compare the raw computational capability of different hardware. Investors track these specifications to understand the performance trends of new chips and how they contribute to the escalating compute requirements of advanced AI models.
Why it matters
TOPS and FLOPS are key performance indicators for AI chips, directly impacting their suitability for different AI tasks. Understanding these metrics helps assess the capabilities of new hardware and its market value.
Reviewed under editorial standardsUpdated September 26, 2026Not investment advice