Snowflake and Databricks are certainly comparable corporations. Whereas every positions itself a bit in a different way, each present knowledge storage, processing and governance in a cloud context. Each are holding buyer conferences this week, and each are in search of methods to assist prospects construct generative AI and different clever functions on prime of the info saved in these platforms.
If that wasn’t clear earlier than, it turned much more obvious this week when Databricks introduced it was buying MosaicML for a cool $1.3 billion. That’s some huge cash for a startup, even a well-capitalized one like Databricks. The transfer got here weeks after the corporate introduced it was releasing Dolly, an open supply LLM, and one other acquisition in AI governance software Okera.
Snowflake introduced final month that it was shopping for Neeva, giving it a search software and a few high-end AI engineering expertise. The corporate additionally purchased Streamlit final 12 months, which lets corporations construct functions from the info saved in Snowflake, and on Wednesday, it introduced a brand new container service and partnership with Nvidia, giving prospects a method to construct generative AI functions and run them on Nvidia GPUs.
All of those strikes (and others) are designed with one factor in thoughts: to make use of the info saved in these companies as gasoline for machine studying fashions, particularly massive language fashions. Each corporations wish to assist prospects reap the benefits of all this knowledge saved on their platforms.
Nvidia’s VP of enterprise computing, Manuvir Das, talking within the context of Wednesday’s partnership announcement with Snowflake, sees the transfer towards extra sensible use of the info as a logical development for Snowflake.
“The truth that Snowflake is now transferring on this subsequent step the place they’re saying, OK, not solely can you retain your knowledge right here and do form of the apparent knowledge processing issues on it, however that is the place the place you may construct all of the functions that drive your organization as a result of your knowledge is true right here. That’s a really highly effective factor,” Das informed TechCrunch+.
Equally, Databricks is more and more seeing itself as a spot the place you cannot solely retailer knowledge and do the assorted knowledge duties related to that, however you may as well be half of an entire knowledge stack, the place you construct functions on prime.
This week’s MosaicML acquisition was a part of this broader technique to put the info to work in an AI context, stated Ray Wang, founder and principal analyst at Constellation Analysis. That’s one thing that was exhausting for Databricks to do, even with Dolly.
“The AI angle is all about making it simple to amass, handle, practice and deploy LLMs with ease,” Wang stated.
Each corporations are clearly transferring exhausting towards AI by way of acquisitions, partnerships and product growth. However what does that imply from a possible income perspective for the way forward for these corporations, one in all which is already public and one which certainly shall be there finally?
Enterprise AI demand isn’t illusory
Databricks and Snowflake are each rising in a short time. The most recent info from Databricks signifies that in its most up-to-date fiscal 12 months, it generated greater than $1 billion in income, rising at greater than 60%. Snowflake’s outcomes are equally spectacular, posting $623.6 million in income in its most up-to-date quarter, up 48% in comparison with the year-ago interval.