Edit

HBv2 series virtual machine overview

  • Article

Caution

This article references CentOS, a Linux distribution that is nearing End Of Life (EOL) status. Please consider your use and plan accordingly. For more information, see the CentOS End Of Life guidance.

Applies to: ✔️ Linux VMs ✔️ Windows VMs ✔️ Flexible scale sets ✔️ Uniform scale sets.

Maximizing high performance compute (HPC) application performance on AMD EPYC requires a thoughtful approach memory locality and process placement. Below we outline the AMD EPYC architecture and our implementation of it on Azure for HPC applications. We use the term pNUMA to refer to a physical NUMA domain, and vNUMA to refer to a virtualized NUMA domain.

Physically, an HBv2-series server is 2 * 64-core EPYC 7V12 CPUs for a total of 128 physical cores. Simultaneous Multithreading (SMT) is disabled on HBv2. These 128 cores are divided into 16 sections (8 per socket), each section containing 8 processor cores. Azure HBv2 servers also run the following AMD BIOS settings:

Nodes per Socket (NPS) = 2
L3 as NUMA = Disabled
NUMA domains within VM OS = 4
C-states = Enabled

As a result, the server boots with 4 NUMA domains (2 per socket) each 32 cores in size. Each NUMA has direct access to 4 channels of physical DRAM operating at 3200 MT/s.

To provide room for the Azure hypervisor to operate without interfering with the VM, we reserve 8 physical cores per server.

VM topology

We reserve these 8 hypervisor host cores symmetrically across both CPU sockets, taking the first 2 cores from specific Core Complex Dies (CCDs) on each NUMA domain, with the remaining cores for the HBv2-series VM. The CCD boundary isn't equivalent to a NUMA boundary. On HBv2, a group of four consecutive (4) CCDs is configured as a NUMA domain, both at the host server level and within a guest VM. Thus, all HBv2 VM sizes expose 4 NUMA domains that appear to an OS and application. 4 uniform NUMA domains, each with different number of cores depending on the specific HBv2 VM size.

Process pinning works on HBv2-series VMs because we expose the underlying silicon as-is to the guest VM. We strongly recommend process pinning for optimal performance and consistency.

Hardware specifications

Hardware Specifications HBv2-series VM
Cores 120 (SMT disabled)
CPU AMD EPYC 7V12
CPU Frequency (non-AVX) ~3.1 GHz (single + all cores)
Memory 4 GB/core (480 GB total)
Local Disk 960 GiB NVMe (block), 480 GB SSD (page file)
Infiniband 200 Gb/s HDR Mellanox ConnectX-6
Network 50 Gb/s Ethernet (40 Gb/s usable) Azure second Gen SmartNIC

Software specifications

Software Specifications HBv2-series VM
Max MPI Job Size 36000 cores (300 VMs in a single virtual machine scale set with singlePlacementGroup=true)
MPI Support HPC-X, Intel MPI, OpenMPI, MVAPICH2, MPICH, Platform MPI
Additional Frameworks UCX, libfabric, PGAS
Azure Storage Support Standard and Premium Disks (maximum 8 disks)
OS Support for SRIOV RDMA CentOS/RHEL 7.9+, Ubuntu 18.04+, SLES 12 SP5+, WinServer 2016+
Orchestrator Support CycleCloud, Batch, AKS; cluster configuration options

Note

Windows Server 2012 R2 is not supported on HBv2 and other VMs with more than 64 (virtual or physical) cores. For more information, see Supported Windows guest operating systems for Hyper-V on Windows Server.

Next steps