Fluence has published The AI Compute Race, a study comparing cloud GPU rental prices for companies that are neither frontier research labs nor hyperscalers. According to the report, hyperscalers price the H100, H200 and B200 models at 1.98 times the median GPU-cloud price, with premiums ranging from 1.4 times for the B200 to 2.8 times for the H100. A one-year commitment reduces the median price by 21%, while a three-year term lowers it by 48%; spot capacity is listed at 53% of the on-demand rate. The report also shows that 17 of the 22 tracked GPU models became more expensive over the past year as demand for open models increased and buyers secured upcoming capacity.
Source: Read the original article
Disclaimer: this content is for information purposes only and is not financial advice. Cryptocurrencies are highly volatile: you may lose all of your capital. Always do your own research. Legal notice
