Snowflake losses widened to $317 million in its second quarter despite robust growth – in part on increased spend on compute for AI features.
“We are incurring GPU-related costs in order to fulfill customer demand for our newer product features” as CEO Sridhar Ramaswamy told analysts.
There were notable positives however in earnings that beat expectations; product revenues rose 30% year-on-year to $829 million – and CEO Ramaswamy hailed a Q2 “hallmarked by innovation and product delivery.”
Icebergs as well as AI (and security) were on the mind for investors, however, as they grilled Ramaswamy on an August 21, 2024 earnings call.
Snowflake: An Iceberg threat?
The market for commodity data lakes is facing disruption by open source software and standards for lakehouse formats like Apache Iceberg.
Positioning here is frantic: Fierce Snowflake rival Databricks agreed to buy Tabular, a company founded by Apache Iceberg’s creators, in June 2024.
(That brought together the creators of the two leading FOSS lakehouse formats – the other being Databricks’/the Linux Foundation’s Delta Lake.)
Snowflake took its own “Iceberg Tables” GA in June 2024, quoting customers like Booking.com’s Chief Data Officer as appreciating the ability to link existing open source Iceberg-based data lakes and Snowflake – which lets the latter be accessed directly from Snowflake via an Iceberg catalogue integration, using its security, governance and sharing controls.
Snowflake Iceberg Tables at 400 customers
Asked about how Snowflake’s Iceberg position was evolving, CFO Mike Scarpelli said: “We have about 400 customers that are using it. We haven't seen customers move storage out of Snowflake… 400 that we know of are actually using Iceberg with new workloads, are trying that out."
He added: “Storage is still running about 11% of our revenue.”
Snowflake’s CEO added: “The acquisition of Tabular, the company, has no impact on the Iceberg project, which is an Apache open source project.
“This has contributors and program committee members from a number of cloud companies, the hyperscalers, yes, but also other companies.
See also: Containerise everything? What Nutanix’s evolution says about the changing face of IT
“We also have members within Snowflake. So we very much intend for this to be an industry standard that we take a pretty significant role in shaping.
"We actually feel that the Tabular acquisition, in many ways, is a vindication of our strategy to bet on Iceberg because that was the format that was truly interoperable. Hopefully, this is the end of the ‘Betamax wars’ – with everybody centering around the one format that has broad support. We will continue to be a key player in this ecosystem to ensure that the format truly serves everybody and moves the industry forward.”
Snowflake customer data breaches: No impact
A series of high-profile breaches of Snowflake customer environments in May did not have a significant impact on earnings, executives emphasised.
Incident response specialists at Mandiant and Crowdstrike said the attacks “targeted campaign directed at users with single-factor authentication; as part of this campaign, threat actors leveraged credentials previously purchased or obtained through infostealing malware” in a June report.
Snowflake CEO Sridhar Ramaswamy admitted that the company needs to be “more proactive about having the security conversation” – with the string of customer breaches caused by poor downstream configurations.
The CEO said “we have them [customers] talk to our CISOs. We have them talk to our security field CTOs, advise them on best practices” – but added “I would say that there's not really been any noticeable effect or delay in things like our ability to… sign up new customers or get existing customers to deploy new projects” (in the wake of the data breaches.)
Snowpark Container Services is GA on AWS
As well as AI product innovations like Snowflake Cortex, the company also took its “Snowpark Container Services” offering GA on AWS and to Public Preview in all Azure regions this August. Snowflake’s Director of Product Jeff Hollan described the product as letting users “build serverless, fully managed apps, data pipelines, and workloads alongside their data in Snowflake” adding in a LinkedIn post that “whether you're training or inferencing data in an ML model, hosting a data app, or creating GenAI solutions, it simplifies the process for everyone to get their code up and running, managed, and scaled next to their data.”
Exploring it at its first release, The Stack described it simply as “what amounts to a managed, secure Kubernetes service that will allow Snowflake’s customers to run containerised applications inside its data platform for the first time: “You can take any container that you have, give it to Snowflake, and we'll run it” as Hollan put it to us at the time.
Speaking at Snowflake’s 2023 Summit, he said: “It’s ‘here's a container, Snowflake; you run it securely with my data.’ Then we do the heavy lifting. We're managing the cluster, we're doing the updates…it comes bundled with logging, a container registry that is securely hosted. Everyone's cluster is entirely run and governed in their Snowflake account.”
%20Iceberg%20alert%3F%20Snowflake%20losses%20widen){kind=link}