The Snowflake/Databricks rivalry heats up

Presented by:

Welcome to Runtime! Today: Snowflake and Databricks hold dueling data conferences, CISOs at public companies have something new to worry about, and this week in enterprise moves.

(Was this email forwarded to you? Sign up here to get Runtime each week.)

Lakes are now battlefields

Most tech rivalries are more collaborative than combative. Apple makes a ton of money by making Google search the default option on the iPhone while competing directly with Android, AWS is a huge reseller of Windows software on its cloud while in a similar position against Microsoft Azure, and nearly every enterprise software vendor builds integrations with fierce rivals to satisfy customer demand.

And then there's Snowflake and Databricks. The two next-generation data management companies have been locked in competition that has turned feisty several times over the last few years, and this week they were vying for the attention of the enterprise tech buyer with overlapping conferences laying out their visions for the future.

Snowflake went first, hosting the Snowflake Summit in Las Vegas starting on Monday.

It kicked off the week by getting one of its bigger generative AI announcements out of the way, announcing a partnership with Nvidia that will allow Snowflake customers to use its NeMo LLM builder against Snowflake data.
Snowflake also launched its own LLM called Document AI, which will let customers use natural-language queries to find needles in their haystack of unstructured Snowflake data.
And in a sign of its reach across enterprise data sets, it released the Native App framework to allow third-party developers to build applications that can run on Snowflake.
Snowflake rose to prominence thanks to an incredibly sticky cloud-native data warehouse that provided fast and easy access to structured data, and is now trying to expand into other parts of the data stack as it matures.

Databricks actually fired the first shot of the week by announcing the $1.3 billion acquisition of MosaicML on Monday, and kicked off its Data & AI Summit in San Francisco with keynotes on Wednesday.

Databricks has focused much more on evangelizing the concept of data lakes and data lakehouses — in which the data lake is a repository for unstructured data and the lakehouse brings order to the chaos — compared to the data warehouse strategy, which sounds a bit musty in 2023.
It launched a service called LakehouseIQ that is similar to Snowflake's Document AI, but also supports developers that want to use generative AI to help with coding and debugging.
Databricks also announced that its Delta Lake open-source project will support competing open-source formats called Iceberg and Hudi, which are backed by Snowflake and several other data industry players.
And customers will be able to work with data across a wide variety of sources — including Snowflake — with the Databricks Lakehouse Federation, introduced in private preview.

At the end of the week, it's clear both rivals are chasing each other's strengths.

Databricks has historically catered to developers and researchers, while Snowflake is a favorite among corporate users in finance and operations.
But Databricks' embrace of Iceberg and Hudi, as well as the Lakehouse Federation, shows a determination to meet enterprise customers where they are.
For its part, Snowflake wants the AI community to take it more seriously as demand for generative AI technology explodes.

The importance of data to the modern enterprise can't be overstated, and Snowflake and Databricks are fighting to handle one of its most important assets.

And given that everyone from major cloud providers to scrappy startups are pushing both companies to compete, businesses looking to upgrade their data-management strategies should have a wealth of options to move forward.

A MESSAGE FROM HASHICORP

Operational cloud maturity is the key to helping enterprises get the most from multi-cloud, slash costs, and maximize ROI with respect to speed, risk, and efficiency. Highly mature organizations are less likely to waste money on avoidable cloud spending, have an easier time dealing with cloud security issues, and better cope with the ongoing shortage of cloud skills. See the third annual State of Cloud Strategy Survey, commissioned by HashiCorp and conducted by Forrester Consulting.

Dark winds

The SolarWinds hack raised about as much attention at the federal government level as any cybersecurity event in recent history. Now the SEC is considering holding SolarWinds executives responsible for its aftermath, which sent chills down the spine of CISOs this week.

Last week CNN reported that the SEC sent a Wells notice to SolarWinds "that it intends to recommend “civil enforcement action” alleging the company broke federal securities laws in its public statements and “internal controls” related to the hack." On Wednesday, Kim Zetter reported that CISOs — a group already used to being the scapegoat — are now really worried about being held liable for incidents they may not have been able to forsee.

“It’s not common for any Wells notice to be sent to a company in relation to cybersecurity,” former federal prosecutor Mark Rasch told Zetter. But this could be the new normal, and will only increase the pressure on security teams at public companies to plug all the holes.

Enterprise moves

Ido Bukspan was named CEO of Pliops, an enterprise storage chip company, after a long stint at Mellanox and Nvidia.

David Aronchick is now CEO of Expanso, a data-processing startup built around the open-source Bacalhau project.

The Runtime roundup

After years of Microsoft playing coy, we finally have a solid number Azure revenue: In June 2022, CEO Sayta Nadella told Microsoft's board that it had recorded $34 billion in Azure revenue over the past fiscal year, which "means Azure’s share of the market was several percentage points smaller than some analyst firms had estimated," according to court documents seen by The Information.

GitHub went down for about 45 minutes in the middle of the U.S. workday, prompting coffee breaks from coast to coast.

Oracle certified its flagship database for Ampere's Arm processors, furthering the relationship between the two companies.

Nvidia's A800 AI processors could also be subject to U.S. export controls in the near future, after the Department of Commerce put sweeping restrictions on high-performance AI chips last year.

OVHcloud reported a 13% jump in revenue during its third quarter, as European customers continue to embrace cloud services while worrying about the global economy.

Microsoft introduced an AI training course for enterprises looking for help getting current employees up to speed on the generative AI boom.