Last week Nvidia announced brand new investments in three companies which it says are “advancing the state of the art of GPU-accelerated data science”, namely, BlazingDB, Graphistry and H2O.ai.
“AI is fundamentally changing how we interact with data,” said Jeff Herbst, vice president of business development at NVIDIA. “These three companies are all rising stars in the data science industry. Their enthusiastic participation in the GOAI open source ecosystem will be instrumental to their ultimate success.”
BlazingDB offers GPU-accelerated analytics on massive data sets, taking SQL into the GPU realm. Their SQL engine is designed to run on a cluster of distributed GPU servers, promising multiple orders-of-maginitue improvement in query execution. They raised $2.9 million from Nvidia GPU ventures and Samsung Next. Database servers have long been the domain of CPU-based servers so this is one to watch.
Graphistry has been featured at GDC for a number of years, showing strange but stunning GPU-accelerated visualization of big data. Half of their team came from UC Berkeley and they bill themselves as “the first visual investigation platform to handle increasing enterprise-scale workloads”. I couldn’t find how much Nvidia invested, and even Crunchbase lists their previous two funding rounds as “undisclosed”.
Rounding out Nvidia’s GPU Ventures latest, well, venture, is a $40 million round of funding in H2O.ai involving Wells Fargo and other venture capital firms. I still can’t figure out what exactly they do, but H2O.ai’s website claims that 8/10 top banks, 7/10 top insurance companies and 4/10 top healthcare companies use their technology. It reminds me of those “9 out of 10 dentists recommend…” ads.
Without getting too conspiratorial (at this stage anyway), keep in mind that In-Q-Tel is involved in a wide range of AI companies, such as Graphistry. If you haven’t heard of it, In-Q-Tel invests and partners with companies to get tech for the CIA and the US Intelligence Community.