Now, you don’t have to choose expensive, large compute instances just to get higher storage performance. With Hyperdisk, you can dynamically scale storage performance and capacity independently to efficiently meet the storage I/O needs of data-intensive workloads such as data analytics and databases. Built on Titanium, Hyperdisk delivers significantly higher levels of performance, flexibility, and efficiency by decoupling storage processing from the virtual machine host. Hyperdisk is Google Cloud’s latest-generation block storage. This collaboration delivers Intel’s 4th Gen Intel Xeon processor and Google’s custom Intel IPU that unlocks new levels of efficiency and performance.” - Suzi Jewett, General Manager - Intel Xeon Products, Intel Corporation Hyperdisk storage “Building on our successful partnership with Google Cloud since 2016, we're proud to collaborate on Google Cloud’s first storage-optimized VM family. Titanium enables Z3 to deliver up to 200 Gbps of fully encrypted networking, 3x faster packet-processing capabilities than prior generation VMs, near-bare-metal performance, integrated maintenance updates for the majority of workloads, and advanced controls for the more sensitive workloads. The end result is better performance, lifecycle management, reliability, and security for your workloads. Z3 VMs are built on Titanium, Google’s system of purpose-built custom silicon, security microcontrollers, and tiered scale-out offloads. You’ll also receive in-place upgrades that preserve your data through planned maintenance events. This allows you to more predictively plan ahead of a disruptive event, while allowing us to deliver more performant and secure infrastructure. You can then schedule the maintenance event at a time of your choosing, or default to the scheduled time. Z3 VMs receive notice from the system several days in advance of a maintenance event. Z3 VMs come with a variety of new infrastructure lifecycle technologies that provide tighter control and specificity around maintenance. We are excited for the incredible performance and value that these new instances will offer our customers.' - Avi Kivity, Co-founder and CTO, ScyllaDB Enhanced maintenance experience

On these new instances, a cluster of just three ScyllaDB nodes can achieve around 2.2M OPS for writes and mixed workloads and around 3.6M OPS for reads. We observed a 23% increase in write throughput, 24% for mixed workloads, and 14% for reads – and that’s with 8 fewer cores (z3-highmem-88 vs n2-highmem-96). 'When we tested ScyllaDB on the new Z3 instances, ScyllaDB exhibited a significant throughput improvement across workloads versus the previous generation of N2 instances. Initially, Z3 comes in two shapes: 176 vCPU with 1.4T of DDR5 memory, and 88 vCPU with 704 GB of DDR5 memory, both with 36TB of next generation Local SSD. Z3 is ideal for horizontal, scale-out databases, flash-optimized databases, data warehouses, and other applications with dense storage requirements. Z3 is also the debut of next generation Local SSD, with up to 3x disk throughput over prior-generation instances and up to 35% lower disk latency. The Z3 machine series brings together the enterprise-grade performance and reliability of the 4th Gen Intel Xeon scalable processor (code-named Sapphire Rapids), Google’s custom Intel Infrastructure Processing Unit (IPU), and the latest generation of Local SSD to Google Compute Engine and Google Kubernetes Engine customers. With an industry-leading 6M 100% random-read and 6M write IOPs, an incredibly dense storage configuration of up to 409 SSD (GB) per vCPU and a highmem configuration (1 vCPU : 8 GB RAM), Z3 VMs provide a consistent performance and maintenance experience with minimal disruptions for storage-dense workloads such as horizontal, scale-out databases and log analytics workloads, allowing you to reduce total cost of ownership by offering more storage capacity for fewer cores. Last week at Google Cloud Next ‘24, we announced the general availability of the Z3 machine series, our first storage-optimized VM family. Storage-dense workloads need consistent performance, high SSD density, and predictable maintenance that preserves SSD data.

Shadow era hacks