"The big worry centers on GPUs, the chips needed to train and run AI models. As new GPUs come out, older ones get less valuable, through obsolescence and wear and tear. Cloud companies must use depreciation to reduce the value of these assets over a period that reflects reality. The faster the depreciation, the bigger the hit to earnings. Investors have begun to worry that GPUs only have useful lives of one or two years,"
""GPUs can profitably run for about 6 years," Stacy Rasgon, a leading chip analyst at Bernstein, wrote in a research report on Monday. "The depreciation accounting of most major hyperscalers is reasonable." Healthy margins The cost of operating a GPU in an AI data center is "very low" compared to market prices for renting GPUs via the cloud. That makes the "contribution margins" of running old GPUs for longer quite high, Rasgon and his fellow analyst at Bernstein noted."
Concerns about an AI bubble have focused on GPU depreciation risk, with some investors fearing GPUs only last one to two years while providers depreciate over five to six years. Rapid GPU architecture updates drive worries about accelerated obsolescence and potential future earnings hits for cloud operators. Bernstein analysis indicates GPUs can profitably run for about six years and that many hyperscalers' depreciation accounting is reasonable. Operating costs for GPUs in AI data centers remain very low compared with cloud rental prices, resulting in high contribution margins for extending the life of older GPUs.
Read at Business Insider
Unable to calculate read time
Collection
[
|
...
]