# From FLOPs to Footprints: The Resource Cost of Artificial Intelligence

arXiv: https://arxiv.org/abs/2512.04142 · HTML: https://mlsanity.com/paper/2512.04142/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 69 references that are in the map.
- First posted on arXiv 3 Dec 2025.

## Builds on (most cited first, 21 of 21)

- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Scaling Laws for Neural Language Models](https://mlsanity.com/paper/2001.08361/) (2020)
- [Measuring Mathematical Problem Solving With the MATH Dataset](https://mlsanity.com/paper/2103.03874/) (2021)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [HellaSwag: Can a Machine Really Finish Your Sentence?](https://mlsanity.com/paper/1905.07830/) (2019)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [Training Compute-Optimal Large Language Models](https://mlsanity.com/paper/2203.15556/) (2022)
- [FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning](https://mlsanity.com/paper/2307.08691/) (2023)
- [Training Deep Nets with Sublinear Memory Cost](https://mlsanity.com/paper/1604.06174/) (2016)
- [Power Hungry Processing: Watts Driving the Cost of AI Deployment?](https://mlsanity.com/paper/2311.16863/) (2023)
- [Liger Kernel: Efficient Triton Kernels for LLM Training](https://mlsanity.com/paper/2410.10989/) (2024)
- [MegaScale: Scaling Large Language Model Training to More Than 10,000 GPUs](https://mlsanity.com/paper/2402.15627/) (2024)
- [Making AI Less "Thirsty": Uncovering and Addressing the Secret Water Footprint of AI Models](https://mlsanity.com/paper/2304.03271/) (2023)
- [A Comprehensive Study on Quantization Techniques for Large Language Models](https://mlsanity.com/paper/2411.02530/) (2024)
- [From Efficiency Gains to Rebound Effects: The Problem of Jevons' Paradox in AI's Polarized Environmental Debate](https://mlsanity.com/paper/2501.16548/) (2025)
- [More than Carbon: Cradle-to-Grave environmental impacts of GenAI training on the Nvidia A100 GPU](https://mlsanity.com/paper/2509.00093/) (2025)
- [Characterizing the Efficiency of Distributed Training: A Power, Performance, and Thermal Perspective](https://mlsanity.com/paper/2509.10371/) (2025)
- [EMLIO: Minimizing I/O Latency and Energy Consumption for Large-Scale AI Training](https://mlsanity.com/paper/2508.11035/) (2025)

## Built on by (most cited first, 2 of 2)

- [The Global Landscape of Environmental AI Regulation: From the Cost of Reasoning to a Right to Green AI](https://mlsanity.com/paper/2603.00068/) (2026)
- [Wasted large language models: A life cycle thinking approach](https://mlsanity.com/paper/2608.17055/) (2026)

## Abstract (arXiv)

As computational demands continue to rise, assessing the environmental footprint of AI requires moving beyond energy and water consumption to include the material demands of specialized hardware. This study quantifies the material footprint of AI training by linking computational workloads to physical hardware needs. The elemental composition of the Nvidia A100 SXM 40 GB graphics processing unit (GPU) was analyzed using inductively coupled plasma optical emission spectroscopy, which identified 32 elements. The results show that AI hardware consists of about 90% heavy metals and only trace amounts of precious metals. The elements copper, iron, tin, silicon, and nickel dominate the GPU composition by mass. In a multi-step methodology, we integrate these measurements with computational throughput per GPU across varying lifespans, accounting for the computational requirements of training specific AI models at different training efficiency regimes. Scenario-based analyses reveal that, depending on Model FLOPs Utilization (MFU) and hardware lifespan, training GPT-4 requires between 1,174 and 8,800 A100 GPUs, corresponding to the extraction and eventual disposal of up to 7 tons of toxic elements. Combined software and hardware optimization strategies can reduce material demands: increasing MFU from 20% to 60% lowers GPU requirements by 67%, while extending lifespan from 1 to 3 years yields comparable savings; implementing both measures together reduces GPU needs by up to 93%. Our findings highlight that incremental performance gains, such as those observed between GPT-3.5 and GPT-4, come at disproportionately high material costs. The study underscores the necessity of incorporating material resource considerations into discussions of AI scalability, emphasizing that future progress in AI must align with principles of resource efficiency and environmental responsibility.