Increase of Performance & Power of AI over Time

Increase of performance of AI over time, using the following three metrics:

• Accuracy: This metric measures how well an AI model can predict or classify the correct output given some input data. Usually expressed as a percentage or fraction of correct predictions over total predictions. Higher accuracy means higher performance and reliability of the model1.

• Time: This metric measures the time an AI model needs to process a given amount of data or perform a given task. Usually expressed as a rate or a time unit, such as seconds, milliseconds, or operations per second. Less time means higher speed, means higher efficiency & scalability of the model2.

• Energy: This metric measures how much power an AI model consumes when running on a given hardware device or platform. Usually expressed as a unit of power or work, such as watts, joules, or kilowatt-hours. Lower energy means lower environmental impact and cost of the model2.