Carmen Li, who most recently ran global data integration for Bloomberg, has started a new startup, Silicon Data. The firm has created a pioneering GPU rental price index. Now, today, that SDH100RT index was first unveiled. Its purpose is to provide an ongoing rental cost index for Nvidia H100 GPUs. Although the index was officially launched today, Silicon Data has been tracking GPU rental costs for several months in advance.
Li also wants the SDH100RT to inspire banks. More specifically, this effort will allow them to invest in a wider array of actors across the artificial intelligence (AI) ecosystem. By providing a clear picture of rental prices, she believes financial institutions can create products that mitigate risks for those already engaged in the AI industry. This new initiative is a manifestation of her larger belief that having access to computing resources will be a major factor determining success in the years to come.
Understanding the SDH100RT Index
The SDH100RT index, which tracks Nvidia H100 GPU rental prices. That will soon change, as it rolls out to cover other AI chips, including those from AMD, and eventually Nvidia’s Blackwell series. This index is especially timely, as it reveals the devastating regional disparities in GPU rental costs. In part, he was reacting to the rental price of an H100 GPU that Li saw in March. It was $5.76 per hour on the East Coast, but on the West Coast it jumped to $6.80 per hour. This regional disparity clearly demonstrates the urgent need for a transparent, equitable pricing model as AI rapidly transforms the industry.
Li’s deep experience with integrating data and her recent forays into the market have opened her eyes to the greater potential. She noted that hyperscaler chips, such as those produced by Amazon Web Services (AWS), have contributed to lowering prices for cloud consumers. These improvements are rapidly lowering the costs of cloud services. This trend is particularly exciting for businesses eager to gain a competitive edge through emerging AI technologies but weighed down by prohibitive operating costs.
“I really believe compute will be the biggest resource for humanity in the next few years.” – Carmen Li
To provide more comprehensive insight, Silicon Data has introduced the SDH100RT index. This new effort helps to equip entrepreneurs and employees alike with the financial literacy necessary to thrive in the AI sector. Li’s vision is much larger than monitoring fare prices. Her goal is to create a system where the private sector has access to the resources they need at a reasonable cost.
Market Trends and Implications
Since last September, Li has been hard at work analyzing and posting her discoveries on her Medium blog. She’s tracked major developments on the GPU rental price front and analyzed what this means for the burgeoning AI industry. Her insights show that price transparency is essential in creating an environment that supports growth and innovation. The creation of the new SDH100RT index is a direct response to this need.
According to Li’s research, the market was highly responsive to all major advances in AI, including the release of DeepSeek. She saw the entire stock market go nuts, right as DeepSeek was coming out. That’s a measure of just how much financial markets are riding on the rapid development of generative AI technologies. Due to this overwhelming demand, it’s imperative to create trustworthy pricing metrics to maintain stable expectations and inform wise investment opportunities.
The added complication of GPU rental prices being highly influenced by geographic location makes things even trickier. H100 rentals East Coast has a $1.04 per mile advantage vs. West Coast. The bottom line Businesses can’t throw everything at a project and expect to make the most money. The SDH100RT index provides critical context in this space. Most importantly, it gives businesses the power to plan better and more thoroughly make the most of their computing resources.
Future Prospects for AI Financing
Carmen Li has bigger plans than just creating one index with Silicon Data. Primarily, she aims to change how financial institutions engage with tech firms in the booming AI space. She thinks banks have a big role to play in developing new financial products specifically for AI firms. They can do this by empowering customers through improved price transparency.
Li’s experience at Bloomberg has equipped her with a keen understanding of data-driven decision-making, which she now brings into her startup. As more agencies and companies look to AI solutions, the availability of robust and reliable pricing information will be critically important.
The SDH100RT index provides a great base case for banks. That’s good for investors looking to get into—or expand their existing operations in—the AI space. By following trends in rental prices and identifying potential risks, financial institutions may find opportunities for growth and collaboration with emerging AI enterprises.