AI startups are finding it harder to access chipmaker Nvidia’s (NVDA) GPUs, as major cloud providers like Microsoft (MSFT), Google (GOOGL), and Amazon (AMZN) prioritize their own internal projects and large customers. According to The Information, smaller companies are being pushed to the back of the line and often face higher prices and long wait times. Even well-funded startups backed by major investors are feeling the pressure, with some founders saying that access to computing power has become one of their biggest challenges. In fact, wait times for GPUs are expected to continue into late 2026.
Claim 30% Off TipRanks
Forget margin or options. Here's how the pros trade AMZNAt the same time, demand for AI tools has surged, making the shortage even worse than in previous years. As a result, cloud providers are raising prices and requiring longer contracts, sometimes forcing startups into expensive, multi-year deals. For example, one company saw its GPU rental costs jump by more than 30% in just a few months, while others struggled to even secure capacity. It also doesn’t help that cloud providers are tightening their rules, such as requiring minimum commitments or limiting access if GPUs are not fully used.
In response, some firms are looking for alternative solutions. For example, venture capital firms are trying to pool their resources in order to help their portfolio companies, while others are considering buying their own hardware instead of renting from cloud providers. However, this approach comes with higher upfront costs and added complexity. Overall, this shortage demonstrates just how critical computing power has become in the AI race, and how access to it is increasingly becoming a major factor in determining which companies will be successful.
Which Cloud Stock Is the Better Buy?
Turning to Wall Street, out of the three cloud stocks mentioned above, analysts think that MSFT stock has the most room to run. In fact, MSFT’s price target of $573.99 per share implies 36.8% upside potential.
