We are excited to announce the preview of the new Azure DHasv8, DHalsv8 and EHasv8-series virtual machines (VMs), powered by 6th Generation AMD EPYC™ (“Venice”) processors. Designed for the next generation of cloud applications, these VMs deliver expanded compute and memory scalability, enhanced storage and networking performance, and flexible configurations to support demanding enterprise, database, and broad range of AI workloads. The preview includes general purpose (DHasv8-series and DHalsv8-series) and memory-optimized (EHasv8-series) VMs, available with and without local disks. These VMs are in preview in the East US Azure region. To request access to the preview, please fill out the request form.
Key enhancements include expanded VM sizes up to 480 cores to support diverse workload requirements; increased memory capacity of up to 3.8 TiB on the EHasv8-series for memory-intensive workloads; and up to 2x higher remote storage IOPS and networking bandwidth on the largest VM sizes compared to the previous generation.
The new VMs are available in three memory (GiB)-to-core configurations, such as 2:1 (DHalsv8-series, DHaldsv8-series), 4:1 (DHasv8-series, DHadsv8-series), and 8:1 (EHasv8-series, EHadsv8-series), giving customers greater flexibility to align infrastructure resources with their workload requirements.
Greater performance and scalability for modern cloud workloads
Purpose-built to meet the growing demands of agentic AI, the new Azure DHasv8-series and EHasv8-series VMs deliver high core density, large memory capacity and bandwidth, enhanced platform I/O, and flexible configurations, with seamless integration into existing AI services. These capabilities help customers run highly concurrent, CPU-intensive AI workloads at scale. Higher core density enables greater parallelism, supporting multiple agent sandboxes, multi-agent orchestration, parallel tool execution, data retrieval, preprocessing, and other CPU-intensive operations across AI pipelines.
Beyond AI, these VMs are optimized for demanding SQL and other database workloads, combining increased memory bandwidth, larger cache, and enhanced networking to accelerate queries, improve transaction concurrency, and boost cache-sensitive application performance. They enable customers to scale transactional databases, analytical processing, and memory-intensive applications with greater performance and efficiency.
These VMs leverage the latest Azure Boost innovations to enhance performance and security. They feature the next-generation Microsoft Azure Network Adapter (MANA), offering stable, forward-compatible networking drivers for Windows and Linux, and support the NVMe protocol for both local and remote disks.
Together, these capabilities provide a scalable foundation for organizations building and deploying increasingly sophisticated agentic AI applications and many diverse workloads.
Get started with the Azure AMD v8 VM preview
During the preview, these new Azure AMD v8 VMs are available for free. However, your Azure subscription will be billed for fees and applicable taxes associated with use of other Azure resources and services, such as disk storage, used in the deployment of these preview VMs.
These VMs support all Azure managed disk types, with storage billed separately. See Azure managed disk types for options and regional availability.
Explore the capabilities of the next generation of Azure AMD based VMs. To request preview access, complete the request form.