SoSoaring Cloud Costs Pose Challenges for AI and Data Analytics Strategiesaring Cloud Costs Pose Challenges for AI and Data Analytics Strategies

Soaring Cloud Costs Pose Challenges for AI and Data Analytics Strategies

The burgeoning fields of cloud computing and generative AI (genAI) have revolutionized data-driven enterprises by providing robust analytics tools and infrastructure. However, the rapid growth in data volumes and the associated rise in cloud costs are beginning to strain business strategies, posing significant financial challenges for companies.

According to the 2024 State of Big Data Analytics report by SQream, a GPU-based big data platform, the cost of cloud analytics is a growing concern. The survey of 300 senior data management professionals from US companies revealed that 71% frequently face unexpected high charges for cloud analytics. Notably, 5% of these companies encounter significant "bill shocks" every month, 25% every two months, and 41% on a quarterly basis.

The report also highlights a troubling trend: 98% of companies experienced failures in machine learning (ML) projects in 2023, primarily due to escalating cloud expenses. Bill shocks often occur when data workflows exceed the capabilities of existing cloud query engines, forcing enterprises to navigate high costs associated with scalable compute power required for large datasets and complex algorithms.

Deborah Leff, Chief Revenue Officer at SQream, noted that companies are increasingly compelled to limit dataset sizes and simplify data complexity to manage costs, which can adversely affect the quality of business insights. High cloud experimentation costs also deter many organizations from initiating AI/ML projects, with poor data preparation and inadequate cleansing contributing to project failures.

The persistent inflation of cloud costs is expected to continue into 2024, prompting enterprises to implement stringent cost-cutting measures. Data from the Bureau of Labor Statistics' Producer Price Index (PPI) shows a month-over-month increase in data processing services, including cloud computing, with a current year-over-year rise of 3.7%.

This cost pressure is forcing many companies to curtail the complexity of their data queries and limit AI-powered projects. Nearly half of the enterprises surveyed (48%) have reduced the complexity of their queries, while 46% have scaled back their AI initiatives due to financial constraints.

Leff also pointed out that the high cost of cloud services is not solely due to vendor pricing. Many businesses fail to thoroughly assess which of their in-house IT assets could benefit from cloud migration. She highlighted that GPU acceleration, often perceived as costly, can actually reduce overall expenses and accelerate processing. For instance, GPU acceleration has allowed organizations like NCBA, a major online bank, to significantly cut down data pipeline cycle times and update marketing models more efficiently.

Leff urged companies to adopt proactive and innovative approaches to manage their data strategies amidst evolving technologies. With generative AI rapidly advancing, she anticipates major shifts in the IT sector over the next two years, urging businesses to stay ahead of these changes to optimize their data and AI investments.