Source URL: https://cloud.google.com/blog/products/databases/c4a-axion-processors-for-alloydb-now-ga/
Source: Cloud Blog
Title: AlloyDB on Axion-powered C4A instances is generally available
Feedly Summary: At Google Cloud Next ’25, we announced the preview of AlloyDB on C4A virtual machines, powered by Google Axion processors, our custom Arm-based CPUs. Today, we’re glad to announce that C4A virtual machines are generally available!
For transactional workloads, leveraging C4A, AlloyDB provides nearly 50% better price-performance compared to N series machines for transactional workloads, and up to 2x better throughput performance compared to Amazon’s Graviton4-based offerings, bringing substantial efficiency gains to your database operations.
Let’s dive into the benefits of running high-performance databases and analytics engines on C4A.
Meeting the demands of modern data workloads
Business-critical applications require infrastructure that delivers high performance, cost-efficiency, high availability, robust security, scalability, and strong data protection. However, the relentless growth of data volumes and increasingly complex processing requirements, coupled with slowing traditional CPU advancements, can present significant hurdles. Furthermore, the rise of sophisticated analytics and vector search for generative AI introduces new challenges to maintaining performance and efficiency. These trends necessitate greater computing power, but existing CPU architectures don’t always align perfectly with database operational needs, potentially leading to performance bottlenecks and inefficiencies.
C4A virtual machines are engineered to address these challenges head-on, bringing compelling advantages to your managed Google Cloud database workloads. They deliver significantly improved price-performance compared to N-series instances, translating directly into more cost-effective database operations. Purpose-built to handle data-intensive workloads that require real-time processing, C4A instances are well-suited for high-performance databases and demanding analytics engines. Integrating Axion-based C4A VMs for AlloyDB underscores our commitment to adding this technology across our database portfolio, for a diverse range of database deployments.
AlloyDB on Axion: Pushing performance boundaries
AlloyDB, our PostgreSQL-compatible database service designed for the most demanding enterprise workloads, sees remarkable gains when running on Axion C4A instances, delivering:
3 million transactions per minute
up to 45% better price-performance than N series VMs for transactional workloads
up to 2x higher throughput and 3x better price-performance than Amazon Aurora PostgreSQL running on Graviton4
Gigaom independently validated performance and price-performance between AlloyDB (C4A) and Amazon Aurora for PostgreSQL, and published a detailed performance evaluation report.
AlloyDB on Axion supports vCPUs from 1vCPU up to 72vCPUs. It also includes a frequently requested intermediate 48 vCPU shape between 32 and 64vCPUs.
In particular, AlloyDB added support for 1 vCPU shape with 8GB of memory with C4A, a cost-effective option for development and sandbox environments. This new shape lowers the entry cost for AlloyDB by 50% compared to the earlier 2vCPU option. The 1 vCPU shape is intended for smaller databases, typically in the tens of gigabytes, and does not come with uptime SLAs, even when configured for high availability.
Availability and getting started
Already, hundreds of customers have adopted C4A machine types, citing increased performance for their applications and reduced costs. They also love the fact that they can lower their costs by 50% using 1vCPU, fueling widespread adoption.. And now that C4A machine types are generally available, customers are excited to leverage them for their production workloads.
Axion-based C4A machines for AlloyDB are now generally available in select Google Cloud regions.
To deploy a new C4A machine for your new instance, or to upgrade your existing instance to C4A, simply choose C4A from the drop-down.
For more technical details, please consult the documentation pages for AlloyDB. For detailed pricing information, please refer to the AlloyDB pricing page. C4A vCPU and memory are priced the same as N2 virtual machines and provide better price-performance. For more information about Google Axion, please refer to Google Axion Processors.
Ready to get started? If you’re new to Google Cloud, sign up for an account via the AlloyDB free trial link. If you’re already a Google Cloud user, head over to the AlloyDB console. Once you’re in the console, click “Create a trial cluster" and we’ll guide you through migrating your PostgreSQL data to a new AlloyDB database instance
AI Summary and Description: Yes
Summary: The text discusses the general availability of Google Cloud’s C4A virtual machines for AlloyDB, highlighting significant performance improvements, cost benefits, and the capacity to handle demanding data workloads. This advancement has implications for professionals focusing on cloud computing security, infrastructure optimization, and performance tuning for database operations.
Detailed Description:
The announcement from Google Cloud regarding the release of C4A virtual machines represents a noteworthy advancement in cloud infrastructure capabilities, particularly for those working with database management and high-performance computing. Key highlights include:
– **Performance Improvements**:
– The C4A instances provide nearly 50% better price-performance for transactional workloads compared to N-series machines.
– They achieve up to 2x higher throughput compared to Amazon Aurora PostgreSQL running on Graviton4.
– **Support for High-Performance Databases**:
– Designed specifically for business-critical applications that require high performance, cost efficiency, and scalability.
– Capable of handling the complexities of modern data workloads and high-volume transactions.
– **Technical Specifications**:
– Supports a range of vCPU configurations from 1 to 72, including a new 1 vCPU shape that is intended for smaller databases and development environments.
– Offers a 50% cost reduction for entry-level database deployment using the 1 vCPU shape.
– **Market Validation**:
– Gigaom has independently validated the performance metrics against competitors like Amazon Aurora, ensuring credibility in the claimed performance gains.
– **Adoption and Availability**:
– Hundreds of customers have already transitioned to C4A machines, citing increased performance and reduced operational costs.
– Available in select Google Cloud regions, the machines are poised for production workloads enhancing the customer experience in database management.
Additional Implications for Security and Compliance Professionals:
– The efficiency and performance metrics discussed may allow companies to optimize their infrastructure security frameworks by reallocating resources, thereby enhancing their security posture without increasing costs.
– The introduction of new virtual machine configurations also raises considerations for compliance and data governance, particularly in relation to workload management and resource handling against regulatory standards.
Overall, the announcement indicates a significant step forward in improving the capabilities of cloud database offerings, presenting valuable insights for professionals within security, compliance, and cloud infrastructure domains.