Beyond the AI buzz at Google Cloud Next 2025 (which Xebia covered last week), important core platform enhancements were announced. Google Cloud introduced updates to make your workloads faster, safer, and simpler to handle. Here are the platform updates you shouldn’t miss.

Compute and infrastructure: more power.

Compute resources remain the fundamental building blocks of Cloud services. Google Cloud Next saw notable improvements across the entire compute platform, including Compute Engine, and Kubernetes Engine.

Next-Gen Virtual Machines

Google Cloud announced significant upgrades to its Virtual Machine lineup, providing more power and options tailored for specific needs. * The general-purpose C-family of VMs has been upgraded with the introduction of C4 (powered by 6th Gen Intel Granite Rapid) and C4D (powered by 5th Gen AMD EPYC processors). These new VMs boast a remarkable 30% performance improvement over the C3D series. * For specialized workloads, Google also announced: * H4D: VMs for High-Performance Computing (HPC). * M4: VMs optimized for in-memory SAP HANA deployments. * Z3: VMs optimized for storage-optimized workloads. * Titanium, Google’s “offload family” received updates for new SSDs, ML adapters, and integration with GPU clusters.

Kubernetes gets smarter (GKE)

Google Kubernetes Engine (GKE) received updates aimed at improving the management, performance, and cost-effectiveness of containerized workloads.

Simplified Management for Large Clusters: Cluster Director for GKE is now generally available, making it easier to manage massive clusters of specialized VMs for AI and HPC workloads, reducing operational complexity.
AI-Aware Load Balancing: The GKE Inference Gateway (in preview) introduces AI model-aware load balancing, promising to significantly cut serving costs and accelerate response times.
Autopilot performance improvements: Serverless GKE Autopilot now features faster pod scheduling, quicker scaling, and improved capacity right-sizing, ensuring your applications are always running optimally.

Infrastructure improvements

Several updates targeted the foundational hardware and management layers:

Managed Instance Groups (MIGs) as Single Entity: An update allows MIGs – collections of VMs managed together for scalability and high availability – to be managed as a single entity. This drastically simplifies the operational management of resilient and dynamically scaling applications, minimizing administrative overhead and potential errors.
Cloud WAN: Enterprises can now use Google’s own super-fast, private global network backbone to connect their sites and data centers. Google is claiming big performance boosts and cost savings compared to a customer-managed WAN solution.
Beefier Interconnect: Cloud Interconnect and cross-cloud interconnect are getting a 4x speed boost to 400 Gbps!
Service extensions for load balancers: Powered by WebAssembly, this feature (now generally available) lets you run custom code at the edge and on load-balancing data paths. Use cases include adding headers, performing redirects, checking for PII, or validating JWT tokens. Cloud CDN support is coming soon.

Storage: Faster data access

Google Cloud’s storage offerings received significant updates focused on speed and efficiency:

Speedy object storage: The new Rapid Storage zonal buckets promise crazy-fast (<1ms) access for latency-sensitive data.
SSD-backed zonal read cache for Cloud Storage: Anywhere Cache lets you create caches in the same zone as your workloads. When you create a cache in a zone, data read requests originating from the zone are processed by the cache instead of the bucket. This means faster data access, but also reducing multi-region data transfer fees. Anywhere cache can be enabled on existing buckets.

Data and Databases

Google Cloud’s data and database services also saw impressive updates:

SQL Server support for DMS: Database Migration Service now supports SQL Server to PostgreSQL migrations for Cloud SQL and AlloyDB. This is supported for both self-managed and cloud-managed SQL server offerings.
Cloud SQL and AlloyDB on C4A: In preview, Cloud SQL and AlloyDB instances are now available on C4A instances, powered by Google’s custom ARM Axion processors. Google claims a nearly 50% improvement in price-performance compared to traditional N-series (x86) machines.
MongoDB compatibility in Firestore: Google expanded Firestore with MongoDB compatibility, which is a highly requested feature. Users can leverage existing MongoDB API, while benefiting from the serverless and highly available features that Firestore offers.
Manage your database fleet with Database Center: This new dashboard provides a unified view of your database fleet, supporting various Google Cloud database products. It offers organization-level insights and helps address compliance, security, data protection, and availability issues.

Conclusion

Google Cloud Next 2025 delivered a wealth of updates beyond the much-discussed AI innovations. These enhancements across compute, storage, and databases underscore Google’s commitment to strengthening its core platform. It’s clear that Google Cloud is investing heavily across the board to provide faster, more secure, and more manageable services.

Google Cloud Next 2025 – Beyond AI