# NVIDIA outlines how AI factories boost return on investment with productivity, durability and flexibility

Digest AI · Hardware & Compute · published 2026-10-01T13:00:49Z

Canonical: https://digestai.news/story/nvidia-outlines-how-ai-factories-boost-return-on-investment-with-produ

## Summary

NVIDIA’s blog post explains how its AI factories maximize return on investment through three key factors: earning capacity, useful life, and demand. Each megawatt-scale factory costs about **$60 million**, and operators prioritize clear ROI before committing capital. NVIDIA’s AI factories are designed to be **productive**—delivering the highest throughput per megawatt and lowest cost per token—**durable**—extending hardware’s earning life beyond five years—and **fungible**, running every type of AI workload (language, vision, biology, robotics) and non-AI tasks like data processing and simulation.

The company highlights **NVIDIA Vera Rubin NVL72** systems, which deliver **over 30x higher throughput per megawatt** than **NVIDIA GB300 NVL72** and up to **45x lower cost per million tokens** for the **DeepSeek V4 Pro** model. NVIDIA’s **CUDA-X libraries** enable running any accelerated workload, while continuous software optimization keeps installed hardware productive for years. The blog also notes that **A100 GPUs**, introduced in **2020**, remain in commercial use six years later, with operators extending depreciation schedules. A **September 2026 Sprout analysis** estimates useful life for an eight-GPU **H100 system** at **five to six years** and for **GB300 NVL72** at **nine to ten years**, based on resale value.

## Key points

- NVIDIA AI factories cost about **$60 million per megawatt** and focus on productivity, durability, and flexibility to maximize ROI
- **Vera Rubin NVL72** systems offer **30x higher throughput per megawatt** and **45x lower token costs** than **GB300 NVL72** for **DeepSeek V4 Pro**
- **CUDA-X libraries** enable running AI and non-AI workloads across generations of NVIDIA GPUs, extending hardware’s useful life

## Why it matters

NVIDIA’s approach reduces costs, extends hardware lifespan, and broadens demand by supporting diverse workloads, making AI infrastructure more efficient and adaptable for operators.

## Sources

1. [Productive, Durable, Fungible: How NVIDIA AI Factories Maximize Return on Investment](https://blogs.nvidia.com/blog/productive-durable-fungible-ai-factories) (NVIDIA Blog, 2026-10-01, primary source)

Part of the developing story: [AI Power Struggle Between Anthropic and Nvidia](https://digestai.news/thread/opinion-amazon-may-benefit-more-than-alphabet-from-anthropics-2-trillion-ipo) (3 stories)

## Cite

Digest AI, "NVIDIA outlines how AI factories boost return on investment with productivity, durability and flexibility", 1 October 2026, https://digestai.news/story/nvidia-outlines-how-ai-factories-boost-return-on-investment-with-produ

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/nvidia-outlines-how-ai-factories-boost-return-on-investment-with-produ.json