Microsoft recently announced the general availability (GA) of new optimized storage Azure Virtual Machines (VMs). These VMs are the Lasv3 and Lsv3 series designed to run workloads that require high throughput and IOPS, including big data applications, SQL and NoSQL databases, distributed file systems, and data analytics engines.
The Lasv3 and Lsv3 series of virtual machines offer faster processors, increased networking, and higher remote disk throughput compared to previous generation storage-optimized virtual machines. More precisely:
Additionally, high-speed, low-latency, directly-mapped LANs Nonvolatile Express Memory (NVMe) storage are features of the new Lasv3 and Lsv3 virtual machines. Additionally, VMs are available in capacities ranging from 8 to 80 vCPUs – and there’s 8 GiB of memory per vCPU and a 1.92 TB NVMe SSD device for 8 vCPUs, with the largest VM sizes offering up to to 19.2TB (10×1.92TB).
At Databricks, we’ve evaluated Azure Lsv3 and Lasv3 VMs for Lakehouse using a comprehensive set of benchmarks that reflect common and critical patterns found in real-world analytics workloads. Local NVMe v3 L-series devices and faster processor options enable Databrick Photonthe native vectorized query engine, to deliver industry-leading price/performance for the most demanding data warehousing workloads.
Note that Microsoft’s competitor in the cloud space, AWS, also provides storage-optimized VMs with Amazon EC2 I3 instances.
Currently, Lsv3 virtual machines are available in over 20 regions. Lasv3 VMs are initially available in US Central, US East, US South Central, and Western Europe regions, with more to follow. And finally, VM pricing details are available on the pricing pages for the Windows and linux.