en.Wedoany.com Reported - Google Cloud has officially launched the C4N virtual machine instances optimized for networking and storage, expanding its Compute Engine product line for data-intensive workloads.

The new instance series targets applications requiring high network throughput or large block storage, covering databases, network appliances, analytics, and some artificial intelligence inference tasks. These machines are based on Google Cloud's Titanium offload architecture, which shifts networking and storage processing to dedicated hardware.

The C4N is equipped with fifth-generation Intel Xeon Scalable processors (codename Emerald Rapids). Instances offer up to 400 Gbps of network bandwidth and up to 95 million packets per second, with block storage throughput reaching 25 GiB/s and up to 1 million input/output operations per second when paired with Hyperdisk Extreme.

Google Cloud positions the C4N as a way to help customers avoid over-provisioning compute resources to meet I/O demands. The series scales bandwidth, packet processing, and storage performance across instance sizes ranging from 2 to 192 virtual CPUs, with memory configurations up to 1.5 TB.

Google Cloud differentiates the C4N from its general-purpose C4 virtual machines by tuning it for predictable performance under sustained networking and storage demands. For smaller instance sizes, the C4N offers 25 Gbps to 50 Gbps of network bandwidth in 2 to 16 vCPU configurations, suitable for customers running I/O-intensive tasks without needing additional compute power. These machines also boost internet egress bandwidth to up to 200 Gbps, with egress packet processing speeds reaching 48 million packets per second.

Storage is a core focus of this launch. The C4N supports the full Hyperdisk portfolio, including Balanced, Balanced High Availability, Extreme, Throughput, and ML options. With Hyperdisk Balanced, instances can achieve up to 20 GiB/s of block storage throughput and nearly 640,000 IOPS.

Google Cloud also reported application-level improvements over C4 machines in internal testing, including up to a 1.5x increase in Nginx web requests per second at typical request sizes, and up to a 45% improvement in MySQL query throughput when data is primarily stored on disk.

Several partners and customers outlined early use cases for the platform in telecommunications, analytics, storage, and high-performance computing. Ericsson linked these instances to mobile core workloads running in public cloud environments. Eric Parsons, Vice President and Head of Ericsson On-Demand, stated: "5G core workloads are inherently network-intensive, requiring high-throughput packet processing and deterministic latency, which standard public cloud instances often struggle to sustain at scale. By leveraging the Google Cloud C4N compute series, we found the ideal engine for Ericsson On-Demand. C4N's architectural focus on network-optimized computing enables our 5G core-as-a-service to achieve unprecedented throughput levels—such as our recent 1 Tbps milestone—while maintaining the carrier-grade reliability our customers expect. This is no longer just about cloud-native; with C4N, we are delivering network-native performance in a public cloud environment."

Teradata indicated that these instances are suitable for intensive production workloads combining analytics, AI, and large datasets. Kevin Dougherty, Senior Director of Product Management for Teradata's Core Platform, stated: "Teradata's autonomous knowledge platform unifies production-grade AI, analytics, and data into an integrated system—providing the context, governance, and performance pillars needed for autonomous AI at scale. Customers rely on Teradata to run mission-critical, highly I/O-intensive workloads where performance and cost control directly determine value. Google Cloud C4N instances are well-suited for these demanding workloads, offering a strong price-to-performance ratio and supporting more efficient, optimized deployments. By leveraging C4N on Google Cloud, Teradata Cloud can help customers accelerate from insight to action—confidently scaling enterprise intelligence and generating greater impact from their data and AI investments."

NetApp linked this launch to storage-intensive AI deployments that require more bandwidth between compute and data services. Pravjit Tiwana, Senior Vice President and General Manager of Cloud Storage and Services at NetApp, stated: "With the next-generation network and storage bandwidth of C4N VMs, Google Cloud NetApp Volumes will unlock new performance levels to support our customers' most demanding AI workloads. By expanding Google Cloud NetApp Volumes support for the C4N VM series through collaboration, Google and NetApp are deepening our partnership to address real customer challenges. Together, we deliver in-place AI and analytics solutions that simplify architecture, maximize performance, and turn data into impact."

Sycomp provided one of the more detailed performance reports, testing storage throughput on the instances. Scott Fadden, Senior HPC Solutions Architect at Sycomp, stated: "Most Compute Engine instances come with a single high-speed network interface. The new C4N doubles the bandwidth potential with two 200 GbE interfaces. This architectural change is significant. It means we can dedicate both networks entirely to storage traffic, providing double the available bandwidth for data-intensive workloads and achieving up to 2x the storage performance of the previous generation. The C4N was just released a few weeks ago and is already active in Sycomp's test environment, ensuring our customers can evaluate the latest GCP capabilities in a timely manner. Google Cloud's published maximum Hyperdisk Balanced performance for the C4N is 20 GiB/s. In our tests, using three storage servers, Sycomp achieved 58.5 GiB/s read speeds and 58.6 GiB/s write speeds; using ten C4N storage servers, we achieved 195 GiB/s read and write speeds—reaching 97% of the theoretical limit without any platform-specific tuning. This is a strong starting point, and there is still measurable room to close the remaining gap through configuration work we can fine-tune. The C4N is not just faster—it changes the price-to-performance equation for storage workloads on Google Cloud."

Intel also commented on the processor and infrastructure design behind the product. Srini Krishna, Intel Fellow for Data Center Products at Intel, stated: "Google Cloud's launch of the C4N highlights how infrastructure innovation and a strong silicon foundation can help customers tackle increasingly data-intensive workloads. With Intel Xeon and custom Infrastructure Processing Units (IPUs), the C4N delivers the performance and efficiency required for demanding network-optimized environments."