An Even Easier-to-Use and More Trusted Platform from Snowflake

Many organizations struggle with the complexity and cost of managing all their data, apps and models to quickly extract actionable insights. They waste valuable time and resources on piecemeal solutions that require stitching together disparate services from multiple cloud vendors — leading to manual configurations and upgrades, clunky governance and security, wasteful downtime and underutilized data.

As we step into the era of enterprise AI, organizations' requirements for a robust platform that is easy to use, connected and trusted for their current and future data and AI needs remain unchanged. This is exactly what Snowflake’s fully managed platform provides. Thousands of customers power their most demanding workloads on Snowflake because we help save valuable time and money with a platform that automates administration, scales with their needs, brings AI to their data and constantly saves them on costs through automatic performance improvements and built-in FinOps. These organizations trust Snowflake to protect their most sensitive data because of our out-of-the-box and unified security, governance, observability and business continuity/disaster recovery expertise across regions and clouds.  

Today, we are making a series of announcements that take all these platform capabilities to the next level.

Intelligent compute innovations, enhanced FinOps and easier cross-account management at scale

At Snowflake, we have prioritized making compute even faster and even easier to use with better price/performance. With Standard Warehouse – Generation 2 (Gen2) (generally available), an updated version of Snowflake’s current Standard Warehouse that has upgraded hardware and additional performance enhancements, Snowflake has delivered 2.1x faster performance for core analytics workloads. Snowflake is also excited to introduce Snowflake Adaptive Compute (in private preview), the next evolution of our compute service that takes away the undifferentiated heavy lifting of making infrastructure choices to save time and money. Adaptive Compute enables ultimate ease of use by automatically selecting the appropriate cluster size(s), the number of clusters, and auto-suspend/resume duration for jobs on your behalf to minimize required configurations. Query routing is even done intelligently to the right-sized clusters without any user action.

Furthermore, Adaptive Compute provides better price/performance not only from leveraging the latest and greatest hardware and performance enhancements from Snowflake but also from queries optimally sharing a pool of compute resources across your account to maximize efficiency. Warehouses created using the Adaptive Compute service are called Adaptive Warehouses, which are in private preview.  

Adaptive Warehouses can be used with not only Snowflake’s existing cost visibility, control and optimization capabilities but even our newest FinOps innovations, including:  

• Cost-Based Anomaly Detection (in public preview) in the Cost Management interface, which Snowflake roles other than ACCOUNTADMIN can now access (generally available), alerts you when abnormal cost spikes occur, enabling users to detect, notify and understand the key contributors to abnormal costs. This was previously available as a Streamlit app, but we’re making it even easier to access by incorporating it directly into the Cost Management interface.

• Tag-Based Budgets (generally available) enables you to now establish spending limits for resources based on the tags they are associated with, simplifying the process of both associating resources to teams or business units and establishing spending limits for those entities.

• Enhanced visibility and optimization of query performance is enabled through capabilities such as Performance Explorer (generally available soon), Query Insights (private preview) and Cost Insights for Large Query Gaps (private preview).

To further set and drive adoption of industry best practices for efficiently operating cloud solutions at scale, Snowflake is excited to join the FinOps Foundation as a Premier Enterprise Member. The FinOps Foundation has historically focused on influencing cost management best practices for cloud service providers (CSPs) but is now starting to work with consumption-based SaaS and PaaS providers, making it the perfect time for Snowflake to be more involved. Snowflake already uses the FinOps Foundation’s best practices in customer engagements through our value engineering team, in ongoing releases of built-in cost visibility, control and optimization capabilities aligned to the FinOps framework and in internal forums for efficiently managing our CSP spend at scale. Membership in the FinOps Foundation is another channel for us to further help our customers maximize the value of Snowflake.

We are also making it easier and more efficient to manage Snowflake organizations at scale through enhanced object management, collaboration, FinOps and security posture management across all the associated accounts in an organization. Users with GLOBALORGADMIN roles can create an organization account (public preview), use SQL to query organization usage views (generally available), and set up organization users and user groups and roles (public preview).

AI-powered, interoperable Snowflake Horizon Catalog to govern and discover data in Snowflake, external sources and Apache Iceberg tables 

Snowflake Horizon Catalog provides out-of-the-box, unified security, governance and discovery across regions and clouds. Horizon Catalog federates across Iceberg REST catalogs through Catalog-linked Databases (generally available soon), which automatically syncs Iceberg objects managed by remote catalogs in Horizon Catalog. Data engineers can act on their entire Iceberg ecosystem from a single pane of glass, ensuring all read and write operations stay in sync and are protected by Horizon Catalog’s leading governance and discovery capabilities. In addition to supporting Iceberg data, Horizon Catalog will provide external data discovery (in private preview soon) for external data in relational databases, dashboards, semantic models and more.

We are also introducing enhancements that make Horizon Catalog even easier to use and more automated, including an updated left navigation in Snowsight that allows you to more quickly find and access “Catalog” and “Governance and Security” capabilities. 

Leveraging AI, Snowflake is enabling you to use natural language to do governance, security and metadata discovery tasks in this new Snowsight experience with Copilot for Horizon Catalog (in private preview soon), helping more users get critical information about data sets faster and with fewer dependencies.

Anomaly detection in Trust Center uses both rule-based (generally available) and AI-based models (in public preview soon) to detect anomalous behavior in your Snowflake accounts so you can save time and improve your organization’s security posture by proactively detecting suspicious access attempts to your accounts. Other continuous monitoring updates include Leaked Password Protection (generally available) to automatically disable user passwords discovered on the dark web and Bad IP Protection (generally available soon) to block access to Snowflake accounts from well-known bad IPs. 

We are also announcing a comprehensive suite of other security and governance enhancements to strengthen overall account protection, including:

• Trust Center extensions (generally available soon) that allow customers to easily build scanners customized to their unique environment and configurations to detect threats sooner. They can also tap into the expertise of Snowflake’s security partners who have built integrations with Trust Center.

• New MFA methods and account security updates such as passkey support (Touch ID, FaceID, Windows Hello, security keys) and authenticator apps (generally available). To help support existing BI apps that don’t yet support MFA login, we are also launching programmatic access tokens (PATs) (generally available) as a drop-in replacement for passwords. This helps completely remove static credentials stored in Snowflake. We are also announcing Workload Identity Federation (generally available soon) which allows applications and services to authenticate securely to Snowflake without managing their own long-lived credentials like passwords. 

• Sensitive data monitoring and reporting (in private preview soon) through automatic detection of sensitive data, tagging and reporting using an interactive governance UI. 

• Synthetic data generation (generally available) that allows you to create an altered copy of production data without exposing any of the original underlying data so you can extract analytical value from sensitive information while staying privacy-compliant. 

• Enhanced Private Link support that enables more regulated customers to securely connect inbound from Google Cloud Platform (GCP) to Snowflake (generally available) and outbound from Snowflake to GCP for External: Access, Functions, Stages and Volumes (generally available).

Out-of-the-box observability 

We are making it easier for you to gain enhanced observability with Snowflake Trail for your infrastructure, pipelines, apps and AI across your Snowflake environment. You can now strengthen your trust in your workloads with more ways to understand and optimize performance through:

• Core observability improvements such as improvements in Event Table ingestion latency and an updated pricing model (generally available) that unlock a more cost-effective solution for customers and make telemetry data ingestion up to 10x faster. Traces for SQL Statements in Event Tables (generally available) enable more detailed performance breakdowns for stored procedures, providing better visibility.

• Enhanced infrastructure observability with Snowpark Container Services (SPCS) platform Event Logs (public preview) to easily understand when infrastructure changes occur for better troubleshooting or identification of performance bottlenecks.

• Enhanced pipeline observability with telemetry support for Snowflake Openflow (generally available) that collects and analyzes operational data (logs and metrics) from systems running on Openflow within Snowflake to simplify monitoring and troubleshooting across systems. Gain detailed monitoring for dbt projects on Snowflake with telemetry support (public preview), which provides comprehensive logs, traces and visualizations to enhance developer productivity and improve data pipeline reliability.

• Enhanced application observability by being able to integrate custom application metrics into Snowflake using SPCS application metrics and traces (generally available) for unified visibility and faster troubleshooting alongside other Snowflake activity. 

• Enhanced AI observability (generally available soon) through the ability to evaluate, debug and optimize your generative AI agents and apps. Measure application performance through systematic evaluations, compare different LLM configurations side-by-side and trace agent execution steps securely within the Snowflake platform. Report on nuanced metrics such as relevance, groundedness and harmfulness for deeper understanding of AI systems. These tools improve transparency, enabling thorough benchmarking to boost accuracy, reduce costs and accelerate the deployment of more trustworthy AI.

Immutable data protection and enterprise-grade business continuity/disaster recovery for your data, apps and AI 

Enhanced business continuity and disaster recovery capabilities from Snowflake make it even easier for companies to prepare for outages, prevent cyber threats and meet regulatory requirements through:

• Snapshots (in public preview soon) that create point-in-time, immutable backups of your Snowflake data and objects. Once created for a specified duration, these Snapshots cannot be modified or deleted, even by administrators. This immutability helps ensure regulatory compliance, bolsters cyberresilience against threats such as ransomware and guarantees the long-term integrity of historical data for auditing or legal purposes.

• Streamlined Disaster Recovery Management with a more intuitive Replication interface (generally available) in Snowsight that allows you to configure failover groups, select databases and objects and schedule replication without writing code. In addition, a simplified Bulk Failover interface (in public preview soon) provides a point-and-click UI in Snowsight to initiate disaster recovery for your entire replicated account faster during an outage. This helps minimize downtime and complexity when switching operations to your secondary account.

• Granular control and performance optimization for Account Replication, including Schema-Level Replication in Failover Groups (generally available) that allows you to replicate only specific schemas so you can save costs by replicating only essential data. Furthermore, underlying Replication service performance enhancements (generally available) reduce overhead and latency for faster, more efficient data synchronization and improved Recovery Point Objectives (RPOs).

Easily extend Snowflake for any use case across any workload, at enterprise scale 

Not seeing a first-party solution in Snowflake for your use case? Any custom solution can be deployed seamlessly and securely with SPCS, the extensibility runtime for Snowflake. SPCS is a fully managed service that can be used to build and run a wide range of production use cases, including advanced data engineering pipelines, custom ML model training and deployment, multimodal inference and full-stack AI and business applications. Customers can develop business logic in any language and leverage open source frameworks, tools and models — or effortlessly lift and shift existing services to run securely and natively within Snowflake. 

At Summit, we’re thrilled to announce greater availability of SPCS with general availability soon in GCP regions, enabling custom workloads in production on any public cloud.

Unlocking new ways to leverage transactional data on the Snowflake platform

Snowflake continues to evolve to support a wider range of data types and workloads. Given the growing importance of being able to run transactional workloads alongside analytical ones on the same platform — especially for modern application development using AI — we are excited to meet more of your needs across the end-to-end data lifecycle with:

• Enterprise-ready managed PostgreSQL in Snowflake through Snowflake Postgres (in development), a fully managed, enterprise-grade PostgreSQL database solution within Snowflake. Snowflake Postgres will bring the world's most popular transactional database to developers on Snowflake, while providing the uncompromising governance, security and operational rigor essential for building and running mission-critical AI apps.

• Unistore expansion to Azure with Hybrid Tables in Azure (private preview), which gives businesses using Azure the ability to access Hybrid Tables, the underlying technology powering Unistore. This enables you to eliminate overhead, streamline governance and security, and act on data faster by bringing together analytical and transactional data in Snowflake.

• Security enhancements for Unistore (private preview) that allow organizations requiring the highest level of security and customer-managed control over encryption keys to use Hybrid Tables, which now support Snowflake’s Tri-Secret Secure (TSS) (private preview).

Learn more

Visit the Snowflake Platform webpage for more information.

Forward Looking Statements

This article contains forward-looking statements, including about our future product offerings, and are not commitments to deliver any product offerings. Actual results and offerings may differ and are subject to known and unknown risk and uncertainties. See our latest 10-Q for more information.