Nvidia H100 AI GPU Lead Times Shrink to Eight Weeks
Nvidia's H100 AI GPU shortages show clear signs of easing as delivery wait times drop from up to eleven months to just eight to twelve weeks. High enterprise demand remains, but increased supply and cloud rental options help bridge the gap.
The severe shortage of Nvidia's H100 AI GPUs shows clear signs of easing as delivery lead times drop significantly. According to Dell Taiwan's General Manager Terence Liao, customers now wait just eight to twelve weeks for their orders, a major improvement over the three to four month wait times seen earlier this year and the eleven month delays experienced six months ago.
Despite the improving supply chain, demand for AI-capable hardware remains incredibly high as businesses prioritize AI servers over general-purpose models. However, Liao notes that shorter procurement times are the main driver of this shift, since AI servers carry exceptionally high costs that make long waits highly impractical for most organizations.
This rapid improvement in availability stems from a combination of factors, including some companies reselling their surplus H100 inventory to offset maintenance costs. Additionally, cloud providers like AWS make it easier for businesses to rent H100 GPUs on demand, which further alleviates the pressure on physical hardware purchases and points to a potential end to the shortage by later this year.