The Lsv3-series of Azure Virtual Machines (Azure VMs) features high-throughput, low latency, directly mapped local NVMe storage. These VMs run on the 3rd Generation Intel® Xeon® Platinum 8370C (Ice Lake) processor in a hyper-threaded configuration. This new processor features an all-core turbo clock speed of 3.5 GHz with Intel® Turbo Boost Technology, Intel® Advanced-Vector Extensions 512 (Intel® AVX-512) and Intel® Deep Learning Boost. The Lsv3-series VMs are available in sizes from 8 to 80 vCPUs. There are 8 GiB of memory allocated per vCPU, and one 1.92TB NVMe SSD device allocated per 8 vCPUs, with up to 19.2TB (10x1.92TB) available on the L80s_v3 size.
Note
The Lsv3-series VMs are optimized to use the local disk on the node attached directly to the VM rather than using durable data disks. This method allows for greater IOPS and throughput for your workloads. The Lsv3, Lasv3, Lsv2, and Ls-series VMs don't support the creation of a host cache to increase the IOPS achievable by durable data disks.
The high throughput and IOPS of the local disk makes the Lsv3-series VMs ideal for NoSQL stores such as Apache Cassandra and MongoDB. These stores replicate data across multiple VMs to achieve persistence in the event of the failure of a single VM.
1Temp disk speed often differs between RR (Random Read) and RW (Random Write) operations. RR operations are typically faster than RW operations. The RW speed is usually slower than the RR speed on series where only the RR speed value is listed.
Storage capacity is shown in units of GiB or 1024^3 bytes. When you compare disks measured in GB (1000^3 bytes) to disks measured in GiB (1024^3) remember that capacity numbers given in GiB may appear smaller. For example, 1023 GiB = 1098.4 GB.
Disk throughput is measured in input/output operations per second (IOPS) and MBps where MBps = 10^6 bytes/sec.
Temp disk: Lsv3-series VMs have a standard SCSI-based temp resource disk for use by the OS paging or swap file (D: on Windows, /dev/sdb on Linux). This disk provides 80 GiB of storage, 4,000 IOPS, and 80 MBps transfer rate for every 8 vCPUs. For example, Standard_L80s_v3 provides 800 GiB at 40000 IOPS and 800 MBPS. This configuration ensures the NVMe drives can be fully dedicated to application use. This disk is ephemeral, and all data is lost on stop or deallocation.
NVMe Disks: NVMe disk throughput can go higher than the specified numbers. However, higher performance isn't guaranteed. Local NVMe disks are ephemeral. Data is lost on these disks if you stop or deallocate your VM.
NVMe Disk encryption Lsv3 VMs created or allocated on or after 1/1/2023 have their local NVMe drives encrypted by default using hardware-based encryption with a Platform-managed key, except for the regions listed below.
Remote (uncached) storage info for each size
Size Name
Max Remote Storage Disks (Qty.)
Uncached Premium SSD Disk IOPS
Uncached Premium SSD Throughput (MB/s)
Uncached Premium SSD Burst1 IOPS
Uncached Premium SSD Burst1 Throughput (MB/s)
Uncached Ultra Disk and Premium SSD v2 IOPS
Uncached Ultra Disk and Premium SSD v2 Throughput (MB/s)
1Some sizes support bursting to temporarily increase disk performance. Burst speeds can be maintained for up to 30 minutes at a time.
Storage capacity is shown in units of GiB or 1024^3 bytes. When you compare disks measured in GB (1000^3 bytes) to disks measured in GiB (1024^3) remember that capacity numbers given in GiB may appear smaller. For example, 1023 GiB = 1098.4 GB.
Disk throughput is measured in input/output operations per second (IOPS) and MBps where MBps = 10^6 bytes/sec.
Data disks can operate in cached or uncached modes. For cached data disk operation, the host cache mode is set to ReadOnly or ReadWrite. For uncached data disk operation, the host cache mode is set to None.
Expected network bandwidth is the maximum aggregated bandwidth allocated per VM type across all NICs, for all destinations. For more information, see Virtual machine network bandwidth
Upper limits aren't guaranteed. Limits offer guidance for selecting the right VM type for the intended application. Actual network performance will depend on several factors including network congestion, application loads, and network settings. For information on optimizing network throughput, see Optimize network throughput for Azure virtual machines.
To achieve the expected network performance on Linux or Windows, you may need to select a specific version or optimize your VM. For more information, see Bandwidth/Throughput testing (NTTTCP).
Accelerator (GPUs, FPGAs, etc.) info for each size