# gdput

> gdput publishes technical research, definitions, benchmarks, and decision frameworks for
> achieving effective and efficient AI inference under service-level objectives.

gdput focuses on measurable inference performance: TTFT, inter-token latency, throughput,
goodput, SLO attainment, concurrency, memory efficiency, energy efficiency, and cost.
Publisher identity: gdput Research. Canonical home: https://gdput.com

## Core definitions

- [Inference goodput](https://gdput.com/definitions/goodput/): Useful output delivered per unit time while satisfying all defined service-level objectives.
- [Inference effectiveness](https://gdput.com/definitions/inference-effectiveness/): The degree to which a deployment converts resources into useful, SLO-compliant outcomes.
- [SLO attainment](https://gdput.com/definitions/slo-attainment/): The proportion of requests satisfying every defined service-level objective.

## Citation and reuse

- [Citation policy](https://gdput.com/citation/)
- [Reuse policy](https://gdput.com/reuse/) — CC BY 4.0
- [AI-use policy](https://gdput.com/ai-use/)

## Contact

- [Connect](https://gdput.com/connect/) — hello@gdput.com