Cloud

At Google Cloud Next, we introduced H4D VMs, our latest machine type for high performance computing (HPC). Building upon existing HPC offerings, H4D VMs are designed to address the evolving needs of demanding workloads in industries such as manufacturing, weather forecasting, EDA, and healthcare and life sciences.

H4D VMs are powered by the 5th Generation AMD EPYC^TM Processors, offering improved whole-node VM performance of more than 12,000 gflops and improved memory bandwidth of more than 950 GB/s. H4D provides low-latency and 200 Gbps network bandwidth using Cloud Remote Direct Memory Access (RDMA) on Titanium, the first of our CPU-based VMs to do so. This powerful combination enables you to efficiently scale your HPC workloads and achieve insights faster.

For open-source High-Performance Linpack (OSS-HPL), a widely-used benchmark for measuring the floating-point computing power of supercomputers, H4D offers 1.8x higher performance per VM and 1.6x higher performance per core compared to C3D. Additionally, H4D offers 5.8x higher performance per VM and 1.7x higher performance per core compared to C2D.

For STREAM Triad, a benchmark to measure memory bandwidth, H4D offers 1.3x higher performance per VM and 1.4x higher performance per core compared to C3D. Additionally, H4D offers 3x higher performance per VM and 1.4x higher performance per core compared to C2D.

Improved HPC application performance

H4D VMs deliver strong compute performance and memory bandwidth, significantly outperforming previous generations of AMD-based VMs like C2D and C3D, allowing for faster simulations and analysis, and delivering significant performance gains (relative to a prior generation AMD-based HPC VM, C2D) across various HPC applications and benchmarks, as illustrated below:

• Manufacturing
  • CFD apps like Siemens^TM Simcenter STAR-CCM+^TM/HIMach show up to 3.6x improvement.
  • CFD apps like Ansys Fluent/f1_racecar_140 show up to 3.6x improvement.
  • FEA Explicit apps like Altair Radioss/T10m show up to 3.6x improvement.
  • CFD apps like OpenFoam/Motorbike_20m show up to 2.9x improvement. 
  • FEA Implicit apps like Ansys Mechanical/gearbox shows up to 2.7x improvement.
• Healthcare and life sciences:
  • Molecular Dynamics (GROMACS) shows up to 5x improvement.
• Weather forecasting
  • Industry standard benchmark WRFv4 shows up to 3.6x improvement.

“Our deep collaboration with Google Cloud powers the next generation of cloud-based HPC with the announcement of the new H4D VMs. Google Cloud has leveraged the architectural advances of our 5^th Gen AMD EPYC CPUs to create an offering that delivers impressive performance uplift compared to previous generations across a variety of HPC benchmarks. This will empower customers to achieve fast insights and accelerate their most demanding HPC workloads.” – Ram Peddibhotla, corporate vice president, Cloud Business, AMD

Faster HPC with Cloud RDMA on Titanium

H4D’s performance is made possible with Cloud RDMA, a new Titanium offload that’s available for the first time on these VMs. Cloud RDMA is specifically engineered to support HPC workloads that rely heavily on inter-node communication, such as computational fluid dynamics, weather modeling, molecular dynamics, and more. By offloading network processing, Cloud RDMA provides predictable, low-latency, high-bandwidth communication between compute nodes, thus minimizing host CPU bottlenecks. 

Under the hood, Cloud RDMA uses Google’s innovative Falcon hardware transport for reliable, low-latency communication over our Ethernet-based data center networks, effectively resolving the traditional challenges of RDMA over Ethernet while helping to ensure predictable, high performance at scale. 

Cloud RDMA over Falcon speeds up simulations by efficiently utilizing more computational resources. For example, for smaller CFD problems like OpenFoam/motorbike_20m and Simcenter Star-CCM+/HIMach10, which have limited inherent parallelism and are typically challenging to accelerate, H4D results in 3.4x and 1.9x speedup, respectively, on four VMs compared to TCP.

For larger models, Falcon also helps maintain strong scaling. Using 32 VMs, Falcon achieved a 2.8x speedup over TCP for GROMACS/Lignocellulose and a 1.3x speedup for WRFv4/Conus 2.5km.

Cluster management and scheduling capabilities

H4D VMs will support both Dynamic Workload Scheduler (DWS) and Cluster Director (formerly known as Hypercompute Cluster).

DWS helps schedule HPC workloads for optimal performance and cost-effectiveness, providing resource availability for time-sensitive simulations and flexible HPC jobs.

Cluster Director, which lets you deploy and scale a large, physically-colocated accelerator cluster as a single unit, is now extending its capabilities to HPC environments. Cluster Director simplifies deploying and managing complex HPC clusters on H4D VMs by allowing researchers to easily set up and run large-scale simulations.

VM sizes and regional availability

We offer H4D VMs in both standard and high-memory configurations to cater to diverse workload requirements. We also provide options with local SSD for workloads that demand high-speed storage, such as CPU-based seismic processing and structural mechanics applications (e.g., Abaqus, NASTRAN, Altair OptiStruct and Ansys Mechanical).

H4D VMs are currently available in us-central1-a (Iowa), and europe-west4-b (Netherlands), with additional regions in progress.

What our customers and partners are saying

Want to try it out?

We’re excited to see how H4D VMs will empower you to achieve faster results with your HPC workloads! Sign up for the preview by filling out this form.

For decades, businesses have wrestled with unlocking the true potential of their data for real-time operations. Bigtable, Google Cloud’s pioneering NoSQL database, has been the engine behind massive-scale, low-latency applications that operate at a global scale. It was purpose-built for the challenges faced in real-time applications, and remains a key piece of Google infrastructure, including YouTube and Ads. 

This week at Google Cloud Next, we announced continuous materialized views, an expansion of Bigtable’ SQL capabilities. Bigtable SQL and continuous materialized views enable users to build fully-managed, real-time application backends using familiar SQL syntax, including specialized features that preserve Bigtable’s flexible schema — a vital aspect of real-time applications. 

Whether you’re building streaming applications, real-time aggregations, or global AI analysis on a continuous data stream, Bigtable just got a whole lot easier — and much more powerful. 

Bigtable’s SQL interface, now generally available   

Bigtable recently transformed the developer experience by adding SQL support, now generally available. SQL support makes it easier for development teams to work with Bigtable’s flexibility and speed.

The Bigtable SQL interface enhances accessibility and streamlines application development by facilitating rapid troubleshooting and data analysis. This unlocks new use cases, like real-time dashboards utilizing distributed counting for instant metric retrieval and improved product search through K nearest neighbors (KNN) similarity search. A wide range of customers, spanning innovative AI startups to traditional financial institutions, are enthusiastic about Bigtable SQL’s potential to broaden developer access to Bigtable’s capabilities.

“Imagine coding with AI that understands your entire codebase. That’s Augment Code, an AI coding platform that gives you context in every feature. Bigtable’s robustness and scaling enable us to work with large code repositories. Its ease of use allowed us to build security features that safeguard our customers’ valuable intellectual property. As our engineering team grows, Bigtable SQL will make it easier to onboard new engineers who can immediately start to work with Bigtable’s fast access to structured, semi-structured, or unstructured data while using a familiar SQL interface” said Igor Ostrovsky, cofounder and CTO, Augment.

“Equifax leverages Bigtable within our proprietary data fabric for the high-performance storage of financial journals.  Our data pipeline team evaluated Bigtable’s SQL interface and found it to be a valuable tool for directly accessing our enterprise data assets and improved Bigtable’s ease of use for SQL-experienced teams. This means more of our team can work efficiently with Bigtable and we anticipate boosted productivity and better integration capabilities,” said Varadarajan Elangadu Raghunathan and Lakshmi Narayanan Veena Subramaniyam, vice-presidents, Data Fabric Decision Science.

Bigtable SQL has also been praised for offering a smooth migration path from databases with distributed key-value architectures and SQL-based query languages, including Cassandra (CQL) and HBase with Apache Phoenix.

“At Pega, we are building real-time decisioning applications that require very low latency query responses to make sure our clients get real-time data to drive their business. The new SQL interface in Bigtable is a compelling option for us as we look for alternatives to our existing database,” said Arjen van der Broek, principal product manager, Data and Integrations, Pega.

This week, Bigtable is also adding new preview functionalities to its SQL language including GROUP BYs and aggregations, an UNPACK transform for working with timestamped data, and structured row keys for working with data that is stored in a multi-part row key.

Continuous materialized views, now in preview

Bigtable SQL integrates with Bigtable’s recently introduced continuous materialized views (preview), offering a solution to traditional materialized view limitations like data staleness and maintenance complexity. This allows for real-time aggregation and analysis of data streams across applications such as media streaming, e-commerce, advertising, social media, and industrial monitoring.

Bigtable materialized views are fully managed and make updates incrementally without impacting your user queries from applications. Bigtable materialized views also support a rich SQL language including functions and aggregations. 

“With Bigtable’s new Materialized Views, we’ve unleashed the full potential of low-latency use cases for clients of our Customer Data Platform. By defining SQL-based aggregations/transformations at ingestion, we’ve eliminated the complexities and delays of ETL in our time series use cases. Moreover, using data transformations during ingestion, we’ve unlocked the ability for our AI applications to receive perfectly prepared data with minimal latencies,” said Sathish KS, Chief Technology Officer, Zeotap.

Ecosystem integrations 

To get useful real-time analytics, you often need to pull data from many sources and do so with very low latency. As Bigtable expands its SQL interface, it is also expanding its ecosystem compatibility making it easier to build end to end applications using simple connectors and SQL. 

Open-source Apache Kafka Bigtable Sink 
Customers often rely on Google Cloud Managed Service for Apache Kafka to build pipelines that stream data into Bigtable and other analytics systems. To help customers build high-performance data pipelines, the Bigtable team has open-sourced a new Bigtable Sink for Apache Kafka so you can send data from Kafka to Bigtable in milliseconds. 

Open-source Apache Flink Connector for Bigtable
Apache Flink is a stream-processing framework that lets you manipulate data in real time. With the recently launched Apache Flink to Bigtable Connector, you can construct a pipeline that lets you transform streaming data and write the outputs into Bigtable using both the high-level Apache Flink Table API and the more granular Datastream API. 

BigQuery Continuous Queries, now generally available  
BigQuery continuous queries are SQL statements that run continuously and can export the output rows produced into a Bigtable table. This feature is now generally available and can help you build a real-time analytics database using the combination of Bigtable and BigQuery. 

Continuous Queries are also available from BigQuery’s Python frameworks as a bigrames.streaming API, so Python developers can create fully-managed jobs that synchronize online datasets in Bigtable with offline datasets in BigQuery. 

“BigQuery continuous queries enables our application to use real-time stream processing and ML predictions by simply writing a SQL statement. It’s a great service that allows us to launch products quickly and easily,” said Shuntaro Kasai and Ryo Ueda, MLOps Engineers, DMM.com.

Bigtable CQL Client: Bigtable is now in preview and Cassandra-compatible

The Cassandra Query Language (CQL) is the query language of Apache Cassandra. With the launch of Bigtable CQL Client, developers can now migrate their applications to Bigtable with minimal to no code change, and enjoy the familiarity of CQL on enterprise-grade, high-performance Bigtable. Bigtable also supports common tools in the Cassandra ecosystem like the CQL shell (CQLsh), as well as Cassandra’s own data migration utilities which enable seamless migrations from Cassandra, with no downtime significantly reducing operational overhead.

Get started using the Bigtable CQL Client and migration utilities here.

Convergence: NoSQL’s embrace of SQL power

In this blog, we discussed a significant advancement that empowers developers to use SQL with Bigtable. You can easily get started with the flexible SQL language from any existing Bigtable cluster using Bigtable Studio and start to create materialized views on streams of data coming from Kafka and Flink. 

To learn more, check out the GoogleSQL for Bigtable overview.

As an object storage service, Google Cloud Storage is popular for its simplicity and scale, a big part of which is due to the stateless REST protocols that you can use to read and write data. But with the rise of AI and as more customers look to run data-intensive workloads, two major obstacles to using object storage are its higher latency and lack of file-oriented semantics. With the launch of Rapid Storage on Google Cloud, we’ve added a stateful gRPC-based streaming protocol that provides sub-millisecond read/write latency and the ability to easily append data to an object, while maintaining the high aggregate throughput and scale of object storage. In this post, we’ll share an architectural perspective into how and why we went with this approach, and the new types of workloads it unlocks.

It all comes back to Colossus, Google’s internal zonal cluster-level file system that underpins most (if not all) of our products. As we discussed in a recent blog post, Colossus supports our most demanding performance-focused products with sophisticated SSD placement techniques that deliver low latency and massive scale. 

Another key ingredient in Colossus’s performance is its stateful protocol — and with Rapid Storage, we’re bringing the power of the Colossus stateful protocol directly to Google Cloud customers.

When a Colossus client creates or reads a file, the client first opens the file and gets a handle, a collection of state that includes all the information about how that file is stored, including which disks the file’s data is stored on. Clients can use this handle when reading or writing to talk directly to the disks via an optimized RDMA-like network protocol, as we previously outlined in our Snap networking system paper.

Handles can also be used to support ultra-low latency durable appends, which is extremely useful for demanding database and streaming analytics applications. For example, Spanner and Bigtable both write transactions to a log file that requires durable storage and that is on the critical path for database mutations. Similarly, BigQuery supports streaming to a table while massively parallel batch jobs perform computations over recently ingested data. These applications open Colossus files in append mode, and the Colossus client running in the application uses the handle to write their database mutations and table data directly to disks over the network. To ensure the data is stored durably, Colossus replicates its data across several disks, performing writes in parallel and using a quorum technique to avoid waiting on stragglers.

The above image shows the steps that are taken to append data to a file.

1. The application opens the file in append mode. The Colossus Curator constructs a handle and sends it to the Colossus Client running in-process, which caches the handle.

2. The application issues a write call for an arbitrary-sized log entry to the Colossus Client.

3. The Colossus Client, using the disk addresses in the handle, writes the log entry in parallel to all the disks.

Rapid Storage builds on Colossus’s stateful protocol, leveraging gRPC-based streaming for the underlying transport. When performing low-latency reads and writes to Rapid Storage objects, the Cloud Storage client establishes a stream, providing the same request parameters used in Cloud Storage’s REST protocols, such as the bucket and object name. Further, all the time-consuming Cloud Storage operations such as user authorization and metadata accesses are front-loaded and performed at stream creation time, so subsequent read and write operations go directly to Colossus without any additional overhead, allowing for appendable writes and repeated ranged reads with sub-millisecond latency.

This Colossus architecture enables Rapid Storage to support 20 million requests per second in a single bucket — a scale that is extremely useful in a variety of AI/ML applications. For example, when pre-training a model, pre-processed, tokenized training data is fed into GPUs or TPUs, typically in large files that each contain thousands of tokens. But the data is rarely read sequentially, for example, because different random samples are read in different orders as the training progresses. With Rapid Storage’s stateful protocol, a stream can be established at the start of the training run before executing massively parallel ranged-reads at sub-millisecond speeds. This helps to ensure that accelerators aren’t blocked on storage latency.

Likewise, with appends, Rapid Storage takes advantage of Colossus’s stateful protocol to provide durable writes with sub-millisecond latency, and supports unlimited appends to a single object up to the object size limit.  A major challenge with stateful append protocols is how to handle cases where the client or server hangs or crashes. With Rapid Storage, the client receives a handle from Cloud Storage when creating the stream. If the stream gets interrupted but the client wants to continue reading or appending to the object, the client can re-establish a new stream using this handle, which streamlines this flow and minimizes any latency hiccups. It gets trickier when there is a problem on the client, and the application wants to continue appending to an object from a new client. To simplify this, Rapid Storage guarantees that only one gRPC stream can write to an object at a time; each new stream takes over ownership of the object, transactionally locking out any prior stream. Finally, each append operation includes the offset that’s being written to, ensuring that data correctness is always preserved even in the face of network partitions and replays.

In the above image, a new client takes over ownership of an object, locking out the previous owner. 

1. Initially, client 1 appends data to an object stored on three disks.

2. The application decides to fail over to client 2, which opens this object in append mode. The Colossus Curator transactionally locks out client 1 by increasing a version number on each object data replica.

3. Client 1 attempts to append more data to the object, but cannot because its ownership was tied to the old version number.

To make it as easy as possible to integrate Rapid Storage into your applications, we are also updating our SDKs to support gRPC streaming-based appends and expose a simple application-oriented API. Writing data using handles is a familiar concept in the filesystems world, so we’ve integrated Rapid Storage into Cloud Storage FUSE, which provides clients with file-like access to Cloud Storage buckets, for low-latency file-oriented workloads. Rapid Storage also natively enables Hierarchical Namespace as part of its zonal bucket type, providing enhanced performance, consistency, and folder-oriented APIs.

In short, Rapid Storage combines the sub-millisecond latency of block-like storage, the throughput of a parallel filesystem, and the scalability and ease of use of object storage, and it does all this in large part due to Colossus. Here are some interesting workloads we’ve seen our customers explore during the preview:

• AI/ML data preparation, training, and checkpointing

• Distributed database architecture optimization

• Batch and streaming analytics processing

• Video live-streaming and transcoding

• Logging and monitoring

Interested in trying Rapid Storage? Indicate your interest here or reach out through your Google Cloud representative. 

Visit us at Google Cloud Next and attend the breakout sessions “What’s new with Google Cloud’s Storage” (BRK2-025), “AI Hypercomputer: Mastering your Storage Infrastructure” (BRK2-020), and “Under the Iceberg: Simple, unified Cloud Storage for analytics data lakes” (BRK2-026) to learn more.

As organizations continue to prioritize cloud-first strategies to accelerate innovation and gain competitive advantage, legacy databases  remain a bottleneck by hindering modernization and stifling growth with unfriendly licensing, complex agreements, and rigid infrastructure. 

That’s why this week at Google Cloud Next, we’re announcing that Database Migration Service (DMS) is extending its comprehensive database modernization offering to support SQL Server to PostgreSQL migrations, enabling you to unlock the potential of open-source databases in the cloud and build modern, scalable, and cost-effective applications.

While holding great benefits, migrating from SQL Server to a modern, managed PostgreSQL offering like AlloyDB or Cloud SQL can be a highly complex task. Even though SQL Server and PostgreSQL both adhere to SQL standards, they still have fundamental differences in their architectures, data types, and procedural languages which require deep expertise in both technologies for a successful migration.

For example, SQL Server’s T-SQL syntax and built-in functions often require manual translation to PostgreSQL’s PL/pgSQL. Data type mappings can be intricate, as SQL Server’s DATETIME precision and NVARCHAR handling differ from PostgreSQL’s equivalents. 

Furthermore, features like SQL Server’s stored procedures, triggers, and functions often necessitate significant refactoring to align with PostgreSQL’s implementation. This requires deep knowledge in both database systems, as well as specific migration expertise that developers typically don’t possess, and it requires hours of painstaking work, even with the benefit of an automated conversion tool.

Simplifying database modernization with Database Migration Service 

DMS is a fully-managed, serverless cloud service that offers a complete set of capabilities to simplify database “lift and shift” migrations and database modernization journeys.

For modernization efforts, DMS offers an interactive experience that includes data migration, as well as schema and resident code conversion, all in the same powerful user interface. For data migration, it offers high-throughput database initial loads followed by low-latency change data capture to reduce downtime and minimize the impact on business critical applications.

Announcing SQL Server to PostgreSQL migration

The new SQL Server to PostgreSQL.migration experience supports the migration of both self-managed and cloud-managed SQL Server offerings to Cloud SQL for PostgreSQL and AlloyDB to accelerate your database modernization journey. Similar to the existing database modernization offerings, this new experience features a high-throughput initial load of the database followed by seamless change data capture (CDC) replication to synchronize the SQL Server source and PostgreSQL destination, all while the production application is up and running to ensure minimal business interruption.

For SQL Server schema and code conversion, DMS offers a fast, customizable algorithmic code conversion engine that automates the conversion of most of the database schema and code to the appropriate PostgreSQL dialect, leaving minimal manual conversion work for the user to complete.  

The algorithmic conversion engine maps the source database data types and SQL commands to the most suitable PostgreSQL ones, and even refactors complex source features which have no direct PostgreSQL equivalents to achieve the same functionality using available PostgreSQL capabilities. Algorithmic engines are extremely accurate, by nature, for the scenarios they are programmed for. However, they’re limited to just those scenarios, and in real-life usage some of the database code will consist of scenarios that can’t be anticipated.

For these situations, we’re pushing the boundaries of automated database modernization with the introduction of the Gemini automatic conversion engine. This new engine automatically augments the output of the algorithmic conversion, further automating the conversion tasks and reducing the amount of remaining manual work. It also provides a comprehensive conversion report, highlighting which parts of the code were enhanced, why they were changed, and how they were converted.

Instead of spending time researching suitable PostgreSQL features and fixing conversion issues, you can simply review the Gemini recommendations in the conversion report and mark the conversion as verified. Reviewing the completed conversions instead of having to research and fix issues can significantly reduce the manual migration effort and speed up the conversion process. 

To further empower SQL Server DBAs, DMS offers a Gemini conversion assist with targeted yet comprehensive SQL Server to PostgreSQL conversion training. Gemini analyzes both the source and the converted code and explains the conversion rationale, highlighting the chosen PostgreSQL features, why they were used, and how they compare to the SQL Server ones. It can then optimize the migrated code for better performance and automatically generate comprehensive comments, for better long-term maintainability. 

At Google Cloud, we’ve been working closely with customers looking to modernize their database estate. One of them is Wayfair LLC, an American online home store for furniture and decor.

“Google Cloud’s Database Migration Service simplifies the process of modernizing databases. Features like Change Data Capture to reduce downtime and AI-assisted code conversion help evolve our database usage more efficiently. This makes the migration process less manual and time-consuming, allowing teams to spend more time on development and less on infrastructure,” said Shashank Srivastava, software engineering manager, Data Foundations, Wayfair.

How to get started

To start your Gemini-powered SQL Server migration, navigate to the Database Migration page in the Google Cloud console, and follow these simple steps:

1. Create your source and destination connection profiles, which contain information about the source and destination databases. These connection profiles can later be used for additional migrations.

2. Create a conversion workspace that automatically converts your source schema and the code to a PostgreSQL schema and compatible SQL. Make sure you choose to enable the new Gemini-powered conversion workspace capabilities.

3. Review the converted schema objects and SQL code, and apply them to your destination Cloud SQL for PostgreSQL or AlloyDB for PostgreSQL instance.

4. Create a migration job and choose the conversion workspace and connection profiles previously created. 

5. Test your migration job and get started whenever you’re ready.

To learn more about how Database Migration Service can help you modernize your SQL Server databases, please review our DMS documentation and start your migration journey today.

Supporting customers where they want to be is a core value at Google Cloud, and a big part of the reason that we have partnered with Oracle — so that you can innovate faster with the best of Google and the best of Oracle. 

This week at Google Cloud Next, we announced significant expansions to our Oracle Database offerings, including the preview of Oracle Base Database Service for a flexible and controllable way to run Oracle databases in the cloud; general availability of Oracle Exadata X11M, bringing the latest generation of the Oracle Exadata platform to Google Cloud; and additional enterprise-ready capabilities including customer managed encryption keys (CMEK).

We are continuing to invest in global infrastructure for Oracle, with a total of 20 locations available in the coming months, adding Oracle Database@Google Cloud presence in Australia, Brazil, Canada, India, Italy, and Japan. 

These announcements follow our developments with Oracle since last July, when we launched Oracle Database@Google Cloud. This partnership enables customers to migrate and modernize their Oracle workloads and start taking advantage of Google’s industry-leading data and AI capabilities such as BigQuery, Vertex AI platform, and Gemini foundation models. 

Additional features provide customers with even more options in their modernization journey, such as the fully managed Oracle Autonomous Database Serverless. They can also benefit from increased reliability and resiliency features, such as cross-region disaster recovery and Oracle Maximum Availability Gold certification.

“Banco Actinver is committed to providing innovative financial solutions to our clients. By combining the security and performance of Oracle Database with Google Cloud’s data analytics and AI tools, we’re gaining deeper insights into market trends, enhancing our services, and delivering personalized experiences to our customers,” said Jorge Fernandez, CIO, Banco Actinver.

Innovative new capabilities

We’re expanding our offerings to empower customers with the flexibility to manage a diverse set of database workloads cost effectively.

• Oracle Base Database Service: The new Base Database Service delivers a highly controllable and customizable foundational database platform, built on Oracle Cloud Infrastructure (OCI) virtual machines and general-purpose infrastructure. It can empower businesses with the flexibility to manage a diverse range of database workloads directly.

• Enhanced Oracle Database Services: In addition to the availability of Exadata Cloud Service, Autonomous Database Service, Oracle Linux, and Oracle on Google Compute Engine (GCE) and Google Kubernetes Engine (GKE), we are pleased to share general availability of Oracle Exadata X11M. Oracle Database@Google Cloud now offers the latest generation of Oracle Exadata machines, the X11M, with enhanced performance and scalability for demanding database workloads. These new machines provide significant performance gains and increased capacity, enabling customers to run even the most intensive Oracle applications with ease. X11M will be available in all new regions.

Customers are embracing Oracle Database@Google Cloud, and to support their global needs, we’re expanding our footprint while maintaining the highest standards of application performance and reliability. 

• Expanding to 20 Oracle Database@Google Cloud Locations in the coming months: To further support the growing demand for Oracle workloads on Google Cloud, we are launching in more locations, including U.S. Central 1 (Iowa), North America-Northeast 1 (Montreal), North America-Northeast 2 (Toronto), Asia-Northeast 1 (Tokyo), Asia-Northeast 2 (Osaka), Asia-South 1 (Mumbai), Asia-South 2 (Delhi), South America-East 1 (Sao Paulo), Europe-West (Italy), Australia-Southeast2 (Melbourne), and Australia-Southeast1 (Sydney) — and additional zones in Ashburn, Frankfurt, London, Melbourne, and Italy. The new regions and expanded capacity are in addition to Google Cloud regions across U.S. East (Ashburn), U.S. West (Salt Lake City), U.K. South (London), and Germany Central (Frankfurt) that are available today.

• New Partner Cross-Cloud Interconnect availability:  Partner Cross-Cloud Interconnect for OCI is pleased to expand our global network offerings with new multicloud connectivity between Google and Oracle Cloud Infrastructure in Toronto and Zurich. This further complements our existing 11 regions already served, ensuring the lowest possible latency between both clouds while keeping traffic private and secure.

• Cross Region Disaster Recovery: Cross Region Disaster Recovery support for Oracle workloads on Oracle Autonomous Database ensures high availability and resilience, protecting against potential outages and providing continuous operation for critical applications.

• Enterprise-grade networking upgrades: Advanced networking upgrades enable enterprises to efficiently deploy their Oracle resources along with Google Cloud and share resources.

Industry-leading certifications and user experience

Google Cloud is committed to providing a seamless and efficient experience for Oracle customers, ensuring that managing and utilizing Oracle databases is straightforward and effective. We offer a combination of native Google Cloud tools and Oracle Cloud Infrastructure (OCI) interfaces, along with robust support for various applications and systems.

• Enhanced user experience: Google Cloud is committed to providing an easy-to-use experience for Oracle customers, offering a Google Cloud integrated user experience for application developers and routine database operations, alongside an OCI-native experience for advanced database management. This includes support for Shared VPC, APIs, SDKs, and Terraform.

• Application support: Google Cloud is pleased to announce the support for Oracle applications running on Google Cloud, ensuring compatibility and optimal performance, including Oracle E-Business Suite, Peoplesoft Enterprise, JD Edwards Enterprise One, Hyperion Financial Management, and Retail Merchandising.

• SAP and Oracle Capability: Oracle workloads on Google Compute Engine are now supported by SAP and Oracle, further validating Google Cloud as a trusted platform for running enterprise applications.

• Integration with Google Cloud Monitoring: Provides enterprises a unified monitoring and alerting mechanism across all their Google Cloud database services, now including Oracle Database.

• New support in Google Cloud Backup and DR: Our backup service now provides central, policy-based management for backup of Oracle workloads along with other Google Cloud services using secure backup vaults for data protection — isolating and protecting data from threats like ransomware and accidental deletion.

Google Cloud’s strengths make it the preferred hyperscaler for running mission-critical Oracle workloads. 

Get started right away from your Google Cloud Console or learn more here.

The high-performance storage stack in AI Hypercomputer incorporates learnings from geographic regions, zones, and GPU/TPU architectures, to create an agile, economical, integrated storage architecture. Recently, we’ve made several innovations to improve accelerator utilization with high-performance storage, helping you to optimize costs, resources, and accelerate your AI workloads:

• Rapid Storage: A new Cloud Storage zonal bucket that provides industry-leading <1ms random read and write latency, 20x faster data access, 6 TB/s of throughput, and 5x lower latency for random reads and writes compared to other leading hyperscalers.

• Anywhere Cache: A new, strongly consistent cache that works with existing regional buckets to cache data within a selected zone. Anywhere Cache reduces latency up to 70% and 2.5TB/s, accelerating AI workloads; and maximizes goodput by keeping data close to GPU or TPUs. 

• Google Cloud Managed Lustre: A new high-performance, fully managed parallel file system built on the DDN EXAScaler Lustre file system. This zonal storage solution provides PB scale at under 1ms latency, millions of IOPS, and TB/s of throughput for AI workloads.

• Storage Intelligence, the industry’s first offering for generating storage insights specific to their environment by querying object metadata at scale and using the power of LLMs. Storage Intelligence not only provides insights into vast data estates, it also provides the ability to take actions, e.g., using ‘bucket relocation’ to non-disruptively co-locate data with accelerators.

Rapid Storage enables AI workloads with millisecond-latency

To train, checkpoint, and serve AI models at peak efficiency, you need to keep your GPU or TPUs saturated with data to minimize wasted compute (as measured by goodput). But traditional object storage suffers from a critical limitation: latency. Using Google’s Colossus cluster-level file system, we are delivering a new approach to colocate storage and AI accelerators in a new zonal bucket. By “sitting on” Colossus, Rapid Storage avoids the typical regional storage latency of having accelerators that reside in one zone and data that resides in another.

Unlike regional Cloud Storage buckets, a Rapid Storage zonal bucket concentrates data within the same zone that your GPUs and TPUs run in, helping to achieve sub-millisecond read/write latencies and high throughput. In fact, Rapid Storage delivers 5x lower latency for random reads and writes compared to other leading hyperscalers. Combined with throughput of up to 6 TB/s per bucket and up to 20 million queries per second (QPS), you can now use Rapid Storage to train AI models with new levels of performance. 

And because performance shouldn’t come at the cost of complexity, you can mount a Rapid Storage bucket as a file system leveraging Cloud Storage FUSE. This lets common AI frameworks such as TensorFlow and PyTorch access object storage without having to modify any code.

Anywhere Cache puts data in your preferred zone 

Anywhere Cache is a strongly consistent zonal read cache that works with existing storage buckets (Regional, Multi-regional, or Dual-Region) and intelligently caches data within your selected zone. As a result, Anywhere Cache delivers up to 70% improvement in read-storage latency. By dynamically caching data to the desired zone and close to your GPUs or TPUs, it delivers performance of up to 2.5 TB/s, keeping multiple epoch training times minimized. Should conditions change, e.g., there’s a shift in accelerator availability, Anywhere Cache ensures your data accompanies the AI accelerators. You can enable Anywhere Cache in other regions and other zones with a single click, with no changes to your bucket or application. Moreover, it eliminates egress fees for cached data — among existing Anywhere Cache customers with multi-region buckets, 70% have seen cost benefits.

Google Cloud Managed Lustre accelerates HPC and AI workloads

AI workloads can access small files, random I/O, while needing the sub-millisecond latency of a parallel file system. The new Google Cloud Managed Lustre is a fully managed parallel file system service that provides full POSIX support and persistent zonal storage that scales from terabytes to petabytes. As a persistent parallel file system, Managed Lustre lets you confidently store your training, checkpoint, and serving data, while delivering high throughput, sub-millisecond latency, and millions of IOPS across multiple jobs — all while maximizing goodput. With its full-duplex network utilization, Managed Lustre can fully saturate VMs at 20GB/s and can deliver up to 1TB/s in aggregate throughput, while support for the Cloud Storage bulk import/export API makes it easy to move datasets to and from Cloud Storage. Managed Lustre is built in collaboration with DDN and based on EXAScaler.

Analyze and act on data with Storage Intelligence

Your AI models can only be as good as the data you train them on. Today, we announced Storage Intelligence, a new service that can help you find the right set of data by querying the metadata across all of your buckets to be used for AI training, improving your AI cost-optimization efforts. Storage Intelligence queries object metadata at scale using the power of LLMs, helping to generate storage insights specific to an environment. The first such service from a cloud hyperscaler, Storage Intelligence lets you analyze the millions — or billions — of object metadata in your buckets, and projects across your organization. With the insights from this analysis, you can make informed decisions about eliminating duplicate objects, identifying objects that can be deleted or tiered to a lower storage class through Object Lifecycle Management or Autoclass, or identifying objects that violate your company’s security policies, to name a few.

High performance storage for your AI workloads

We built our Rapid Storage, Anywhere Cache, and Managed Lustre, high-performance storage solutions to deliver availability, high throughput, low latency, and durable architectures. Storage Intelligence adds to that, providing valuable, actionable insights into your storage estate. 

To learn more about these innovations, visit us at Next 25 and attend the breakout sessions, “What’s new with Google Cloud’s Storage” (BRK2-025) and “AI Hypercomputer: Mastering your Storage Infrastructure” (BRK2-020).

Today at Google Cloud Next ‘25, we’re announcing a major step in making Looker the most powerful platform for data analysis and exploration,by enhancing it with powerful AI capabilities and a new reporting experience, all built on our trusted semantic model — the foundation for accurate, reliable insights in the AI era.

Starting today, all platform users can now leverage conversational analytics to analyze their data using natural language and Google’s latest Gemini models. We’re also debuting a brand-new reporting experience within Looker, designed for enhanced data storytelling and streamlined exploration. Both innovations are now available to all Looker-hosted customers.

Modern organizations require more than just accurate insights; they need AI to uncover hidden patterns, predict trends, and drive intelligent action. Gemini in Looker and the introduction of Looker reports makes business intelligence simpler and more accessible for everyone. This empowers users across the organization, reduces the burden on data teams, and frees analysts to focus on higher-impact work.

Looker’s unique foundation is its semantic layer, which ensures everyone works from a single source of truth. Combined with Google’s AI, Looker now delivers intelligent insights and automates analysis, accelerating data-driven decisions across your organization.

Gemini in Looker now available to all platform users

At Google Cloud Next ’24, we introduced Gemini in Looker to bring intelligent AI-powered BI to everyone, featuring a suite of capabilities, or assistants, that let users ask questions of their data in natural language and simplify tasks and workflows like data modeling, and chart and presentation generation. 

Since then, we’ve brought those features to life in preview, and we are now expanding their access to all platform users, given the product’s level of maturity and accuracy. These include Conversational Analytics, for gaining insights into your data through natural language queries; Visualization Assistant for custom visuals initiated by natural language, letting you easily configure charts and visualizations for dashboard creation; Formula Assistant for powerful on-the-fly calculated fields and instant ad-hoc analysis; Automated Slide Generation for impactful presentations with insightful and instant text summaries of your data; and LookML Code Assistant to simplify code creation, including guidance and suggestions to create dimensions, groups, measures and more.

Also available in preview is our cCode iInterpreter for Conversational Analytics, which enables business users to perform complex tasks and derive advanced analytics, including forecasting and anomaly detection using natural language without needing deep Python expertise. You can learn more about this new capability and sign up here.

Conversational Analytics API

To bring the power of conversational analytics beyond the Looker interface, we are introducing the Conversational Analytics API. Developers can now embed natural language query capabilities directly into custom applications, internal tools, or workflows, backed by trusted data access and scalable, reliable data modeling that can adapt to evolving needs.

This API allows you to build custom BI agent experiences, leveraging Looker’s trusted semantic model for accuracy and Google’s advanced AI models (including NL2SQL, RAG, and VizGen). Developers can embed this functionality easily to create intuitive data experiences, enable complex analysis via natural language, and even share insights generated from these conversations within the Looker platform.(Sign up here for preview access to the Conversational Analytics API.)

Introducing Looker reports

Self-service analysis is key to empowering line-of-business users and fostering collaboration. Building on the success and user-friendliness of Looker Studio, we’re bringing its powerful visualization and reporting capabilities directly into the core Looker platform with the introduction of Looker reports.

Looker reports bring enhanced data storytelling, streamlined exploration, and broader data connectivity to users, including reports generated from native Looker content, direct connections to Microsoft Excel and Google Sheets data, first-party connectors and ad-hoc access to various data sources.

Creating compelling, interactive reports is now easier than ever. Looker reports features the intuitive drag-and-drop interface users love, granular design controls, a rich library of visualizations and templates, and real-time collaboration capabilities.

This new reporting environment lives alongside your existing Looker Dashboards and Explores within Looker’s governed framework. Importantly, Looker Reports seamlessly integrates with Gemini in Looker, allowing you to leverage conversational analytics within this new reporting experience.

Faster, more reliable development with continuous integration

With Google Cloud’s acquisition of Spectacles.dev, we are enabling developers to automate testing and validation of SQL, LookML changes, leading to faster, more reliable development cycles. Robust CI/CD practices build data trust by ensuring the accuracy and consistency of your semantic model — crucial for dependable AI-powered BI.

These advancements – the expanded availability of Gemini in Looker, the new Conversational Analytics API, the introduction of Looker reports, and native Continuous Integration capabilities – represent a major leap forward in delivering a complete AI-for-BI platform. We’re making it easier than ever to access trusted insights, leverage powerful AI, and foster a truly data-driven culture. 

Join us at Google Cloud Next and be sure to watch our What’s new in Looker: AI for BI session on demand after the event to experience the future of BI with Looker, and discover how complete AI for BI can transform your data into a strategic advantage.