Run AI study reveals less than 50 percent of AI models make it to production
If you build it, they might not run it, at least as far as AI models are concerned. Run AI, an AI resource management solutions provider, published the 2023 State of AI Infrastructure Survey, which reveals that less than half of AI models made it to the production stage in over 88 percent of the companies. The survey gathered data from 450 industry professionals across the United States and Western Europe.
Interestingly, the study reports that 54 percent of US respondents and 43 percent of European participants feel infrastructure and compute to be the primary challenges in AI development. Last year, 61 percent of the participants reported data being the main issue with AI development. This shift is seen as an opportunity for edge infrastructure companies to develop an interoperable hardware and software stack.
“Organizations must shift their focus from solely acquiring more data to ensuring they have the proper infrastructure in place to effectively process and utilize it,” said Omri Geller, CEO of Run AI.
The 2023 State of AI Infrastructure Survey is the second report from Run AI to provide new insights into organizational challenges in AI development and the opportunity for solutions providers to tackle them.
The study also shows that building complex and advanced AI applications demands many GPUs to handle heavy workloads. The survey indicates that almost 50 percent of the companies with more than 100 GPUs have reliance on multiple third-party companies.
“Despite being on the cloud, organizations are still facing limitations with unlocking the full potential of their data,” Geller added. “This highlights the reality that cloud hasn’t delivered on its on-demand promise and the importance of building a robust and scalable infrastructure.”
Last year, Run AI partnered with Comet, a provider of MLOps platform for machine learning. This collaboration sought to simplify machine learning projects for data scientists, researchers and IT teams. Additionally, enterprise clients could leverage this integrated edge solution to obtain actionable intelligence.
Article Topics
DevOps | edge AI | model development | Run AI
Comments