This post was originally published on Pure Storage
AI is like any other technology: The more you use it, the better and more helpful it becomes, but also, typically, the more expensive it becomes. Oftentimes, companies fail to consider what they’ll do when their AI projects grow to the point where their total cost of ownership (TCO) starts to greatly exceed the value of the project.
AI projects can and usually do start small and from pretty much anywhere. Data scientists can work from a laptop, workstation, cloud resources, or from resources within powerful servers and storage in a data center.
The challenge comes when you scale your AI project. The initial resources for your AI computing may not be able to handle your bigger, more scaled-out AI efforts. At some point, your production-level AI will need more horsepower and a more reliable infrastructure. But DIY solutions can be daunting and cloud AI solutions can be expensive (think ongoing rental costs of cloud compute, networking, and storage).
Here are some TCO considerations as you scale AI from exploratory pilot phases to production.
The Allure, and Danger, of ‘Big AI’
Everyone’s familiar with the term “Big Data.” Now there’s “Big AI.”
AI is growing very quickly and is becoming
— Read the rest of this post, which was originally published on Pure Storage.
