vSphere 8 Update 1 announced! Welcome aboard, OKTA!

Last year marked the debut of vSphere 8, VMware’s virtualization platform for both traditional and next-generation applications. Today, VMware announced the release of vSphere 8 Update 1, which delivers increased operational efficiency for administrators, enhanced performance for high-end AI/ML workloads, and bolstered security throughout the environment. vSphere 8 Update 1 is expected to be initially available later this fiscal quarter ending in April, with general availability following suit in alignment with our updated release model.

vSphere Configuration Profiles

In vSphere 8, VMware previewed the next generation of vSphere cluster configuration management. With the release of vSphere 8 Update 1, this feature, now named vSphere Configuration Profiles, is fully supported and no longer in Technology Preview.

Sphere Configuration Profiles is a new capability in vSphere 8.0 that allows administrators to manage host configurations at the cluster level. This feature allows administrators to define the desired configuration as a JSON document at the cluster level, verify that hosts comply with this configuration, and remediate non-compliant hosts to ensure conformity.

^{Source: https://core.vmware.com/resource/whats-new-vsphere-8-update-1#sec25426-sub1}

Okta Identity Federation for vCenter

In today’s security landscape, identity management and multifactor authentication are crucial components. The latest release of vSphere, vSphere 8 Update 1, introduces support for cloud-based identity providers in vCenter, including the widely used Okta service. With this new capability, vSphere administrators can leverage modern identity management features for enhanced security and streamlined operations.

ESXi Quick Boot for Secure Systems

The concept of Quick Boot was first introduced in vSphere 6.7 and has proven to be incredibly useful. Who wouldn’t want to reboot their system faster? For vSAN customers with large nodes, longer reboot times than the rebuild timeout can cause issues since a rebuild increases cluster load and wear on flash devices.

With TPM 2.0-enabled hosts, a secure boot and attestation process is undertaken to validate host configuration, which is a reliable method to detect malware and misconfigurations. vSphere 8 Update 1 now offers compatibility with Quick Boot, eliminating the need to choose between the two features. This allows users to benefit from both features, providing improved security and faster boot times.

^{Source: https://core.vmware.com/resource/whats-new-vsphere-8-update-1#sec25430-sub2}

vSphere Fault Tolerance with vTPM

VMware vSphere Fault Tolerance (FT) is a powerful feature that ensures the uninterrupted availability of applications running on virtual machines with up to four virtual CPUs. By creating a live shadow instance of the primary virtual machine, vSphere FT mirrors all operations to provide seamless failover in the event of a hardware outage, thus eliminating downtime and preventing data loss. Once failover is triggered, vSphere FT creates a new, secondary virtual machine that provides ongoing protection to the application, delivering continuous availability and ensuring that critical operations can continue without interruption.

^{Source: https://core.vmware.com/resource/whats-new-vsphere-8-update-1#sec25430-sub3}

Heterogenous vGPU Profiles on the same GPU

vSphere 8 Update 1 now allows for the assignment of different application types, such as VDI (Virtual Desktop Infrastructure), compute, and graphics, to vGPUs on the same physical GPU. This new capability enables administrators to improve GPU utilization, reduce workload fragmentation within GPUs, and ultimately lower costs. By leveraging this feature, admins can ensure that GPU resources are used more effectively, allowing for better performance and more efficient use of resources.

^{Source: https://blogs.vmware.com/vsphere/2023/03/announcing-vsphere-8-update-1.htm}

Support for NVIDIA NVSwitch:

In vSphere 8 Update 1, administrators can experience a significant boost in the performance of AI/ML applications thanks to support for NVIDIA NVSwitch. This technology allows for the connection of up to 8 GPUs per host and up to 8 GPUs to the same virtual machine, enabling administrators to build high-performance, large-scale GPU clusters for AI/ML workloads. By leveraging this feature, admins can achieve faster performance and more efficient use of resources, helping to drive innovation and accelerate insights for their organizations.

^{Source: https://core.vmware.com/resource/whats-new-vsphere-8-update-1#sec25433-sub1}

Host different GPU workloads on a single GPU

In previous versions of vSphere, every NVIDIA vGPU workload running on an ESXi host was required to utilize the same vGPU profile type and GPU memory capacity.

However, with the release of vSphere 8 U1, it is now possible to assign distinct vGPU profile types to NVIDIA vGPUs. While the GPU memory sizes of all profiles must still be the same, this update provides more flexibility for sharing GPU resources among various workloads. For example, a diagram depicting three virtual machines with unique vGPU profile types “B, C, and Q” and GPU memory capacity of “8GB” demonstrates this capability.

Source: https://core.vmware.com/resource/whats-new-vsphere-8-update-1#sec25426-sub3

NVIDIA provides different vGPU profile types depending on the workload.

Profile type A is designed for application streaming or session-based solutions
Profile type B is designed for VDI applications
Profile type C is designed for compute-intensive applications like machine learning
Profile type Q is designed for graphically intensive applications

For more on vGPU Profile Types see the Nvidia Grid vGPU user guide.

Check out these helpful resources: