Google’s first Axion processor, the C4A, is now GA with Titanium SSD. Google Cloud is excited to announce currently that C4A virtual machines with Titanium SSD, which Google specifically built for cloud applications requiring real-time data processing, low latency, and high-throughput storage performance, are now generally available. In comparison to SSDs from earlier generations, Titanium SSDs on C4A virtual machines (VMs) offer storage performance of up to 2.4M random read IOPS, up to 10.4 GiB/s of read throughput, and up to 35% shorter access latency.
Titanium SSDs free up CPU resources by offloading local storage operations, improving storage security and speed. A Titanium Offload Processor connects the host CPU to a system of custom silicon, hardware, and software on-host and throughout Google Cloud’s data centres. This are the first generation of Google SSDs integrated with Titanium, a system that improves application performance by offloading networking, storage, and management from the host CPU.
Based on Google Axion Processors, the C4A family of virtual machine instances offers up to 60% more energy efficiency and up to 65% better price-performance compared to equivalent current-generation x86-based instances. For a variety of Arm-compatible general-purpose workloads, including high-performance databases, analytics engines, search, and workloads that benefit from caching and local storage capacity, C4A and Titanium SSDs work together to provide industry-leading price-performance.
Up to 72 virtual CPUs, 576 GB of memory, and 6 TB of local storage are available in the C4A with Titanium SSDs in two different configurations: Standard (with 4 GB of memory per vCPU) and High-memory (with 8 GB of memory per vCPU). With conventional bandwidth of up to 50 Gbps and Tier networking for high-traffic applications up to 100 Gbps, C4A provides the connection and storage performance required by business workloads. Additionally, Google Cloud’s most recent Balanced and Extreme Hyperdisk storage is supported by C4A instances, offering scalable, high-performance storage with up to 350k IOPS and 5 GB/s throughput per virtual machine.
Customers of Google Cloud can utilise C4A with Titanium SSD in Batch, Dataproc, Google Kubernetes Engine (GKE), Compute Engine, and other applications. Additionally, C4A virtual machines (VMs) are now previewing in Dataflow, supporting Cloud SQL and AlloyDB, with additional services to follow.
What Google Cloud’s customers and partners are saying
Couchbase Capella Columnar was developed specifically to enable AI-driven applications and speed up intricate analytical queries for real-time insights. Capella Columnar offers unmatched price-performance, ultra-low latency, and scalable computational power for operational and analytical applications when it runs on Google Axion C4A instances with Titanium SSDs. Google Cloud is excited to use Capella Columnar on Google Axion processors to assist businesses in providing exceptional customer experiences.
Google Cloud is thrilled to launch Axion-based C4A virtual machines with Titanium SSDs on Databricks on Google Cloud, which will allow us to provide Google Cloud’s clients with ever-greater price-performance and efficiency improvements. Customers will be able to get more out of their investments in data warehousing and artificial intelligence on the Databricks Data Intelligence Platform with the optimizations provided by Google’s most recent Axion computing and Hyperdisk storage solutions.
“With Google Cloud’s AI-powered observability, security, and search solutions on Google Cloud, Elastic is dedicated to helping customers drive innovation and cost-efficiency.” In Google Cloud’s tests, Google Axion-based C4A virtual machines with Titanium SSDs offered throughput that was up to 40% higher than that of earlier generations of virtual machines. Google Cloud is excited to launch Elastic Cloud’s Google Cloud C4A virtual machines with local Titanium SSDs.
Titanium SSD
This specially made local SSD disc offers improved SSD security, speed, and administration with the utilization of Titanium I/O offload processing. Compared to the previous generation of local SSDs, titanium delivers lower latency, better storage IOPS, and throughput. Titanium SSD is used for local SSD storage in the general-purpose C4A machine series and the storage-optimized Z3 machine series. Within their host server, titanium SSD discs are directly connected to the compute instances.
The first general-purpose, Axion-based instance with Titanium SSD, C4A virtual machines, is now widely accessible. Google created C4A virtual machines with Titanium SSDs specifically for cloud applications that need low latency and high throughput storage performance for real-time data processing. Titanium SSDs free up CPU resources by offloading local storage operations, improving storage security and speed.
