VMware vSphere

Virtual GPU Software R460 for VMware vSphere Release Notes

Release information for all users of NVIDIA virtual GPU software and hardware on VMware vSphere.

These Release Notes summarize current status, information on validated platforms, and known issues with NVIDIA vGPU software and associated hardware on VMware vSphere.

1.1. NVIDIA vGPU Software Driver Versions

Each release in this release family of NVIDIA vGPU software includes a specific version of the NVIDIA Virtual GPU Manager, NVIDIA Windows driver, and NVIDIA Linux driver.

For details of which VMware vSphere releases are supported, see Hypervisor Software Releases.

1.2. Compatibility Requirements for the NVIDIA vGPU Manager and Guest VM Driver

The releases of the NVIDIA vGPU Manager and guest VM drivers that you install must be compatible. If you install an incompatible guest VM driver release for the release of the vGPU Manager that you are using, the NVIDIA vGPU fails to load.
See VM running an incompatible NVIDIA vGPU guest driver fails to initialize vGPU when booted.

Compatible NVIDIA vGPU Manager and Guest VM Driver Releases

The following combinations of NVIDIA vGPU Manager and guest VM driver releases are compatible with each other.

• NVIDIA vGPU Manager with guest VM drivers from the same release
• NVIDIA vGPU Manager with guest VM drivers from different releases within the same major release branch
• NVIDIA vGPU Manager from a later major release branch with guest VM drivers from the previous branch

The following table lists the specific software releases that are compatible with the components in the NVIDIA vGPU software 12 major release branch.

Incompatible NVIDIA vGPU Manager and Guest VM Driver Releases

The following combinations of NVIDIA vGPU Manager and guest VM driver releases are incompatible with each other.

• NVIDIA vGPU Manager from a later major release branch with guest VM drivers from two or more major releases before the release of the vGPU Manager
• NVIDIA vGPU Manager from an earlier major release branch with guest VM drivers from a later branch

The following table lists the specific software releases that are incompatible with the components in the NVIDIA vGPU software 12 major release branch.

1.3. Updates in Release 12.4

New Features in Release 12.4

• Security updates - see Security Bulletin: NVIDIA GPU Display Driver - October 2021, which is posted shortly after the release date of this software and is listed on the NVIDIA Product Security page
• Miscellaneous bug fixes

1.4. Updates in Release 12.3

New Features in Release 12.3

• Security updates - see Security Bulletin: NVIDIA GPU Display Driver - July 2021
• Miscellaneous bug fixes

Hardware and Software Support Introduced in Release 12.3

• Support for VMware Horizon 2106 (8.3)

Feature Support Withdrawn in Release 12.3

• Red Hat Enterprise Linux 8.3 is no longer supported as a guest OS.

1.5. Updates in Release 12.2

New Features in Release 12.2

• Support for unified memory with NVIDIA vGPU
• Support for GPUDirect technology remote direct memory access (RDMA)
• Security updates - see Security Bulletin: NVIDIA GPU Display Driver - April 2021
• Miscellaneous bug fixes

Hardware and Software Support Introduced in Release 12.2

• Support for the following GPUs:
  • NVIDIA^® A10
  • NVIDIA RTX A5000
• Support for Red Hat Enterprise Linux 8.4 as a guest OS
• Support for VMware Horizon 2103 (8.2)
• Support for NVIDIA GPU Operator with the following combinations of hypervisor, container platform, and guest OS:
  • VMware vSphere 7.0, Red Hat OpenShift 4.7, and Red Hat CoreOS 4.7
  • VMware vSphere 7.0 Update 2 and later compatible updates, VMware Tanzu Kubernetes Grid, and Ubuntu 20.04

1.6. Updates in Release 12.1

New Features in Release 12.1

• Support for NVIDIA vGPU on GPUs based on the NVIDIA Ampere architecture on VMware vSphere ESXi 7.0 U2 and later compatible updates
• Miscellaneous bug fixes

1.7. Updates in Release 12.0

New Features in Release 12.0

• Support for NVIDIA^® GPU Operator
• Miscellaneous bug fixes

Hardware and Software Support Introduced in Release 12.0

• Support for the following GPUs:
  • NVIDIA A100 HGX 80GB
  • NVIDIA A40
  • NVIDIA RTX A6000
• Support for Windows 10 October 2020 Update (20H2) as a guest OS

  Windows 10 May 2021 Update (21H1), which is a bug fix release for Windows 10 October 2020 Update (20H2), is also supported.

• Support for Red Hat CoreOS 4.7 as a guest OS
• Support for SuSE Linux Enterprise Server 15 SP2
• Support for VMware Horizon 2012 (8.1)

This release family of NVIDIA vGPU software provides support for several NVIDIA GPUs on validated server hardware platforms, VMware vSphere hypervisor software versions, and guest operating systems. It also supports the version of NVIDIA CUDA Toolkit that is compatible with R460 drivers.

2.1. Supported NVIDIA GPUs and Validated Server Platforms

This release of NVIDIA vGPU software provides support for the following NVIDIA GPUs on VMware vSphere, running on validated server hardware platforms:

• GPUs based on the NVIDIA Maxwell™ graphic architecture:
  • Tesla M6 (NVIDIA Virtual Compute Server (vCS) is not supported.)
  • Tesla M10 (vCS is not supported.)
  • Tesla M60 (vCS is not supported.)
• GPUs based on the NVIDIA Pascal™ architecture:
  • Tesla P4
  • Tesla P6
  • Tesla P40
  • Tesla P100 PCIe 16 GB (vSGA, vMotion with vGPU, and suspend-resume with vGPU are not supported.)
  • Tesla P100 SXM2 16 GB (vSGA, vMotion with vGPU, and suspend-resume with vGPU are not supported.)
  • Tesla P100 PCIe 12GB (vSGA, vMotion with vGPU, and suspend-resume with vGPU are not supported.)
• GPUs based on the NVIDIA Volta architecture:
  • Tesla V100 SXM2 (vSGA is not supported.)
  • Tesla V100 SXM2 32GB (vSGA is not supported.)
  • Tesla V100 PCIe (vSGA is not supported.)
  • Tesla V100 PCIe 32GB (vSGA is not supported.)
  • Tesla V100S PCIe 32GB (vSGA is not supported.)
  • Tesla V100 FHHL (vSGA is not supported.)
• GPUs based on the NVIDIA Turing™ architecture:
  • Tesla T4 (vSGA is not supported.)
  • Quadro RTX 6000 in displayless mode (vSGA is not supported.)
  • Quadro RTX 6000 passive in displayless mode (vSGA is not supported.)
  • Quadro RTX 8000 in displayless mode (vSGA is not supported.)
  • Quadro RTX 8000 passive in displayless mode (vSGA is not supported.)

  In displayless mode, local physical display connectors are disabled.

• GPUs based on the NVIDIA Ampere architecture:
  • NVIDIA A100 HGX 80GB (supports only compute workloads on Linux with NVIDIA Virtual Compute Server and GPU pass through; graphics acceleration is not supported)
  • NVIDIA A100 PCIe 40GB (supports only compute workloads on Linux with NVIDIA Virtual Compute Server and GPU pass through; graphics acceleration is not supported)
  • NVIDIA A100 HGX 40GB (supports only compute workloads on Linux with NVIDIA Virtual Compute Server and GPU pass through; graphics acceleration is not supported)
  • NVIDIA A40 in displayless mode (vSGA is not supported.)
  • Since 12.2: NVIDIA A10 (vSGA is not supported.)
  • NVIDIA RTX A6000 in displayless mode (vSGA is not supported.)
  • Since 12.2: NVIDIA RTX A5000 in displayless mode (vSGA is not supported.)

  In displayless mode, local physical display connectors are disabled.

For a list of validated server platforms, refer to NVIDIA GRID Certified Servers.

2.1.1. Switching the Mode of a GPU that Supports Multiple Display Modes

Some GPUs support displayless and display-enabled modes but must be used in NVIDIA vGPU software deployments in displayless mode.

The GPUs listed in the following table support multiple display modes. As shown in the table, some GPUs are supplied from the factory in displayless mode, but other GPUs are supplied in a display-enabled mode.

A GPU that is supplied from the factory in displayless mode, such as the NVIDIA A40 GPU, might be in a display-enabled mode if its mode has previously been changed.

To change the mode of a GPU that supports multiple display modes, use the displaymodeselector tool, which you can request from the NVIDIA Display Mode Selector Tool page on the NVIDIA Developer website.

2.1.2. Switching the Mode of a Tesla M60 or M6 GPU

Tesla M60 and M6 GPUs support compute mode and graphics mode. NVIDIA vGPU requires GPUs that support both modes to operate in graphics mode.

Recent Tesla M60 GPUs and M6 GPUs are supplied in graphics mode. However, your GPU might be in compute mode if it is an older Tesla M60 GPU or M6 GPU or if its mode has previously been changed.

To configure the mode of Tesla M60 and M6 GPUs, use the gpumodeswitch tool provided with NVIDIA vGPU software releases. If you are unsure which mode your GPU is in, use the gpumodeswitch tool to find out the mode.

For more information, refer to gpumodeswitch User Guide.

2.1.3. Requirements for Using C-Series vCS vGPUs

Because C-Series vCS vGPUs have large BAR memory settings, using these vGPUs has some restrictions on VMware ESXi.

• The guest OS must be a 64-bit OS.
• 64-bit MMIO and EFI boot must be enabled for the VM.
• The guest OS must be able to be installed in EFI boot mode.
• The VM’s MMIO space must be increased to 64 GB as explained in VMware Knowledge Base Article: VMware vSphere VMDirectPath I/O: Requirements for Platforms and Devices (2142307).
• Because the VM’s MMIO space must be increased to 64 GB, vCS requires ESXi 6.0 Update 3 and later, or ESXi 6.5 and later.

2.1.4. Requirements for Using vGPU on GPUs Requiring 64 GB or More of MMIO Space with Large-Memory VMs

Some GPUs require 64 GB or more of MMIO space. When a vGPU on a GPU that requires 64 GB or more of MMIO space is assigned to a VM with 32 GB or more of memory on ESXi , the VM’s MMIO space must be increased to the amount of MMIO space that the GPU requires.

For more information, see:

• VMware Knowledge Base Article: VMware vSphere VMDirectPath I/O: Requirements for Platforms and Devices (2142307)
• VMs with 32 GB or more of RAM fail to boot with GPUs requiring 64 GB or more of MMIO space

With ESXi 6.7 or later, no extra configuration is needed.

The following table lists the GPUs that require 64 GB or more of MMIO space and the amount of MMIO space that each GPU requires.

2.1.5. Requirements for Using GPUs Requiring Large MMIO Space in Pass-Through Mode

• The following GPUs require 32 GB of MMIO space in pass-through mode:
  • Tesla V100 (all 16GB variants)
  • Tesla P100 (all variants)
  • Tesla P6
• The following GPUs require 64 GB of MMIO space in pass-through mode.
  • Quadro RTX 8000 passive
  • Quadro RTX 6000 passive
  • Tesla V100 (all 32GB variants)
  • Tesla P40
• Pass through of GPUs with large BAR memory settings has some restrictions on VMware ESXi:
  • The guest OS must be a 64-bit OS.
  • 64-bit MMIO must be enabled for the VM.
  • If the total BAR1 memory exceeds 256 Mbytes, EFI boot must be enabled for the VM.
    Note:

    To determine the total BAR1 memory, run nvidia-smi -q on the host.

  • The guest OS must be able to be installed in EFI boot mode.
  • The Tesla V100, Tesla P100, and Tesla P6 require ESXi 6.0 Update 1 and later, or ESXi 6.5 and later.
  • Because it requires 64 GB of MMIO space, the Tesla P40 requires ESXi 6.0 Update 3 and later, or ESXi 6.5 and later.

    As a result, the VM’s MMIO space must be increased to 64 GB as explained in VMware Knowledge Base Article: VMware vSphere VMDirectPath I/O: Requirements for Platforms and Devices (2142307).

2.1.6. Linux Only: Error Messages for Misconfigured GPUs Requiring Large MMIO Space

In a Linux VM, if the requirements for using C-Series vCS vGPUs or GPUs requiring large MMIO space in pass-through mode are not met, the following error messages are written to the VM's dmesg log during installation of the NVIDIA vGPU software graphics driver:

            

            
NVRM: BAR1 is 0M @ 0x0 (PCI:0000:02:02.0)
[  90.823015] NVRM: The system BIOS may have misconfigured your GPU.
[  90.823019] nvidia: probe of 0000:02:02.0 failed with error -1
[  90.823031] NVRM: The NVIDIA probe routine failed for 1 device(s).
        

2.2. Hypervisor Software Releases

Supported VMware vSphere Hypervisor (ESXi) Releases

Supported Management Software and Virtual Desktop Software Releases

2.3. Guest OS Support

NVIDIA vGPU software supports several Windows releases and Linux distributions as a guest OS. The supported guest operating systems depend on the hypervisor software version.

2.3.1. Windows Guest OS Support

2.3.2. Linux Guest OS Support

2.4. NVIDIA CUDA Toolkit Version Support

The releases in this release family of NVIDIA vGPU software support NVIDIA CUDA Toolkit 11.2.

For more information about NVIDIA CUDA Toolkit, see CUDA Toolkit 11.2 Documentation.

2.5. vGPU Migration Support

vGPU migration, which includes vMotion and suspend-resume, is supported only on a subset of supported GPUs, VMware vSphere Hypervisor (ESXi) releases, and guest operating systems.

Supported GPUs

• Tesla M6
• Tesla M10
• Tesla M60
• Tesla P4
• Tesla P6
• Tesla P40
• Tesla V100 SXM2
• Tesla V100 SXM2 32GB
• Tesla V100 PCIe
• Tesla V100 PCIe 32GB
• Tesla V100S PCIe 32GB
• Tesla V100 FHHL
• Tesla T4
• Quadro RTX 6000
• Quadro RTX 6000 passive
• Quadro RTX 8000
• Quadro RTX 8000 passive
• NVIDIA A10
• NVIDIA A40
• NVIDIA A100 HGX 80GB
• NVIDIA A100 PCIe 40GB
• NVIDIA A100 HGX 40GB
• NVIDIA RTX A5000
• NVIDIA RTX A6000

Supported VMware vSphere Hypervisor (ESXi) Releases

• Release 7.0 and compatible updates support vMotion with vGPU and suspend-resume with vGPU.
• Release 6.7 U1 and compatible updates support vMotion with vGPU and suspend-resume with vGPU.
• Release 6.7 supports only suspend-resume with vGPU.
• Releases earlier than 6.7 do not support any form of vGPU migration.

Supported Guest OS Releases

Windows and Linux.

Known Issues with vGPU Migration Support

2.6. Multiple vGPU Support

To support applications and workloads that are compute or graphics intensive, multiple vGPUs can be added to a single VM. The assignment of more than one vGPU to a VM is supported only on a subset of vGPUs and VMware vSphere Hypervisor (ESXi) releases.

Supported vGPUs

Only Q-series and C-series vGPUs that are allocated all of the physical GPU's frame buffer are supported.

Maximum vGPUs per VM

NVIDIA vGPU software supports up to a maximum of four vGPUs per VM on VMware vSphere Hypervisor (ESXi).

Supported Hypervisor Releases

VMware vSphere Hypervisor (ESXi) 7.0 and 6.7 U3 and later compatible updates only.

If you upgraded to VMware vSphere 6.7 Update 3 from an earlier version and are using VMs that were created with that version, change the VM compatibility to vSphere 6.7 Update 2 and later. For details, see Virtual Machine Compatibility in the VMware documentation.

2.7. Peer-to-Peer CUDA Transfers over NVLink Support

Peer-to-peer CUDA transfers enable device memory between vGPUs on different GPUs that are assigned to the same VM to be accessed from within the CUDA kernels. NVLink is a high-bandwidth interconnect that enables fast communication between such vGPUs. Peer-to-Peer CUDA transfers over NVLink are supported only on a subset of vGPUs, VMware vSphere Hypervisor (ESXi) releases, and guest OS releases.

Supported vGPUs

Only Q-series and C-series time-sliced vGPUs that are allocated all of the physical GPU's frame buffer on physical GPUs that support NVLink are supported.

Supported Hypervisor Releases

Peer-to-Peer CUDA Transfers over NVLink are supported on all hypervisor releases that support the assignment of more than one vGPU to a VM. For details, see Multiple vGPU Support.

Supported Guest OS Releases

Linux only. Peer-to-Peer CUDA Transfers over NVLink are not supported on Windows.

Limitations

• Only direct connections are supported. NVSwitch is not supported.
• Only time-sliced vGPUs are supported. MIG-backed vGPUs are not supported.
• PCIe is not supported.
• SLI is not supported.

2.8. GPUDirect Technology Support

GPUDirect^® technology remote direct memory access (RDMA) enables network devices to directly access vGPU frame buffer, bypassing CPU host memory altogether. GPUDirect technology is supported only on a subset of vGPUs and guest OS releases.

Supported vGPUs

Only C-series vGPUs that are allocated all of the physical GPU's frame buffer on physical GPUs based on the NVIDIA Ampere architecture are supported. Both time-sliced and MIG-backed vGPUs that meet these requirements are supported.

Supported Guest OS Releases

Linux only. GPUDirect technology is not supported on Windows.

Supported Network Interface Cards

GPUDirect technology RDMA is supported on the following network interface cards:

• Mellanox Connect-X^® 6 SmartNIC
• Mellanox Connect-X 5 Ethernet adapter card

Limitations

Only GPUDirect technology RDMA is supported. GPUDirect technology storage is not supported.

2.9. Since 12.2: Unified Memory Support

Unified memory is a single memory address space that is accessible from any CPU or GPU in a system. It creates a pool of managed memory that is shared between the CPU and GPU to provide a simple way to allocate and access data that can be used by code running on any CPU or GPU in the system. Unified memory is supported only on a subset of vGPUs and guest OS releases.

Supported vGPUs

Only Q-series and C-series time-sliced vGPUs that are allocated all of the physical GPU's frame buffer on physical GPUs that support unified memory are supported.

Supported Guest OS Releases

Linux only. Unified memory is not supported on Windows.

Limitations

• Only time-sliced vGPUs are supported. MIG-backed vGPUs are not supported.
• When unified memory is enabled for a VM, vGPU migration is disabled for the VM.

2.10. NVIDIA GPU Operator Support

NVIDIA GPU Operator simplifies the deployment of NVIDIA vGPU software with software container platforms on immutable operating systems. An immutable operating system does not allow the installation of the NVIDIA vGPU software graphics driver directly on the operating system. NVIDIA GPU Operator is supported only on specific combinations of VMware vSphere Hypervisor (ESXi) release, container platform, and guest OS release.

Known product limitations for this release of NVIDIA vGPU software are described in the following sections.

3.1. NVENC does not support resolutions greater than 4096×4096

Description

The NVIDIA hardware-based H.264/HEVC video encoder (NVENC) does not support resolutions greater than 4096×4096. This restriction applies to all NVIDIA GPU architectures and is imposed by the GPU encoder hardware itself, not by NVIDIA vGPU software. The maximum supported resolution for each encoding scheme is listed in the documentation for NVIDIA Video Codec SDK. This limitation affects any remoting tool where H.264 encoding is used with a resolution greater than 4096×4096. Most supported remoting tools fall back to software encoding in such scenarios.

Workaround

Use H.265 encoding. H.265 is more efficient than H.264 encoding and has a maximum resolution of 8192×8192.

If you are using Citrix Virtual Apps and Desktops, you can switch to using H.264 hardware encoding with the Use video codec for compression Citrix graphics policy set to to Actively Changing Regions. This policy setting encodes only actively changing regions of the screen (for example, a window in which a video is playing). Provided that the number of pixels along any edge of the actively changing region does not exceed 4096, H.264 encoding is offloaded to the NVENC hardware encoder.

3.2. Issues occur when the channels allocated to a vGPU are exhausted

Description

Issues occur when the channels allocated to a vGPU are exhausted and the guest VM to which the vGPU is assigned fails to allocate a channel to the vGPU. A physical GPU has a fixed number of channels and the number of channels allocated to each vGPU is inversely proportional to the maximum number of vGPUs allowed on the physical GPU.

When the channels allocated to a vGPU are exhausted and the guest VM fails to allocate a channel, the following errors are reported on the hypervisor host or in an NVIDIA bug report:

            

            
Jun 26 08:01:25 srvxen06f vgpu-3[14276]: error: vmiop_log: (0x0): Guest attempted to allocate channel above its max channel limit 0xfb
Jun 26 08:01:25 srvxen06f vgpu-3[14276]: error: vmiop_log: (0x0): VGPU message 6 failed, result code: 0x1a
Jun 26 08:01:25 srvxen06f vgpu-3[14276]: error: vmiop_log: (0x0):         0xc1d004a1, 0xff0e0000, 0xff0400fb, 0xc36f,
Jun 26 08:01:25 srvxen06f vgpu-3[14276]: error: vmiop_log: (0x0):         0x1, 0xff1fe314, 0xff1fe038, 0x100b6f000, 0x1000,
Jun 26 08:01:25 srvxen06f vgpu-3[14276]: error: vmiop_log: (0x0):         0x80000000, 0xff0e0200, 0x0, 0x0, (Not logged),
Jun 26 08:01:25 srvxen06f vgpu-3[14276]: error: vmiop_log: (0x0):         0x1, 0x0
Jun 26 08:01:25 srvxen06f vgpu-3[14276]: error: vmiop_log: (0x0): , 0x0
        

Workaround

Use a vGPU type with more frame buffer, thereby reducing the maximum number of vGPUs allowed on the physical GPU. As a result, the number of channels allocated to each vGPU is increased.

3.3. Total frame buffer for vGPUs is less than the total frame buffer on the physical GPU

Some of the physical GPU's frame buffer is used by the hypervisor on behalf of the VM for allocations that the guest OS would otherwise have made in its own frame buffer. The frame buffer used by the hypervisor is not available for vGPUs on the physical GPU. In NVIDIA vGPU deployments, frame buffer for the guest OS is reserved in advance, whereas in bare-metal deployments, frame buffer for the guest OS is reserved on the basis of the runtime needs of applications.

If error-correcting code (ECC) memory is enabled on a physical GPU that does not have HBM2 memory, the amount of frame buffer that is usable by vGPUs is further reduced. All types of vGPU are affected, not just vGPUs that support ECC memory.

On all GPUs that support ECC memory and, therefore, dynamic page retirement, additional frame buffer is allocated for dynamic page retirement. The amount that is allocated is inversely proportional to the maximum number of vGPUs per physical GPU. All GPUs that support ECC memory are affected, even GPUs that have HBM2 memory or for which ECC memory is disabled.

The approximate amount of frame buffer that NVIDIA vGPU software reserves can be calculated from the following formula:

max-reserved-fb = vgpu-profile-size-in-mb÷16 + 16 + ecc-adjustments + page-retirement-allocation + compression-adjustment

max-reserved-fb

The maximum total amount of reserved frame buffer in Mbytes that is not available for vGPUs.

vgpu-profile-size-in-mb

The amount of frame buffer in Mbytes allocated to a single vGPU. This amount depends on the vGPU type. For example, for the T4-16Q vGPU type, vgpu-profile-size-in-mb is 16384.

ecc-adjustments

The amount of frame buffer in Mbytes that is not usable by vGPUs when ECC is enabled on a physical GPU that does not have HBM2 memory.

• If ECC is enabled on a physical GPU that does not have HBM2 memory ecc-adjustments is fb-without-ecc/16, which is equivalent to 64 Mbytes for every Gbyte of frame buffer assigned to the vGPU. fb-without-ecc is total amount of frame buffer with ECC disabled.
• If ECC is disabled or the GPU has HBM2 memory, ecc-adjustments is 0.

page-retirement-allocation

The amount of frame buffer in Mbytes that is reserved for dynamic page retirement.

• On GPUs based on the NVIDIA Maxwell GPU architecture, page-retirement-allocation = 4÷max-vgpus-per-gpu.
• On GPUs based on NVIDIA GPU architectures after the Maxwell architecture, page-retirement-allocation = 128÷max-vgpus-per-gpu

max-vgpus-per-gpu

The maximum number of vGPUs that can be created simultaneously on a physical GPU. This number varies according to the vGPU type. For example, for the T4-16Q vGPU type, max-vgpus-per-gpu is 1.

compression-adjustment

The amount of frame buffer in Mbytes that is reserved for the higher compression overhead in vGPU types with 12 Gbytes or more of frame buffer on GPUs based on the Turing architecture.

compression-adjustment depends on the vGPU type as shown in the following table.

vGPU Type	Compression Adjustment (MB)
T4-16Q T4-16C T4-16A	28
RTX6000-12Q RTX6000-12C RTX6000-12A	32
RTX6000-24Q RTX6000-24C RTX6000-24A	104
RTX6000P-12Q RTX6000P-12C RTX6000P-12A	32
RTX6000P-24Q RTX6000P-24C RTX6000P-24A	104
RTX8000-12Q RTX8000-12C RTX8000-12A	32
RTX8000-16Q RTX8000-16C RTX8000-16A	64
RTX8000-24Q RTX8000-24C RTX8000-24A	96
RTX8000-48Q RTX8000-48C RTX8000-48A	238
RTX8000P-12Q RTX8000P-12C RTX8000P-12A	32
RTX8000P-16Q RTX8000P-16C RTX8000P-16A	64
RTX8000P-24Q RTX8000P-24C RTX8000P-24A	96
RTX8000P-48Q RTX8000P-48C RTX8000P-48A	238

For all other vGPU types, compression-adjustment is 0.

3.4. Issues may occur with graphics-intensive OpenCL applications on vGPU types with limited frame buffer

Description

Issues may occur when graphics-intensive OpenCL applications are used with vGPU types that have limited frame buffer. These issues occur when the applications demand more frame buffer than is allocated to the vGPU.

For example, these issues may occur with the Adobe Photoshop and LuxMark OpenCL Benchmark applications:

• When the image resolution and size are changed in Adobe Photoshop, a program error may occur or Photoshop may display a message about a problem with the graphics hardware and a suggestion to disable OpenCL.
• When the LuxMark OpenCL Benchmark application is run, XID error 31 may occur.

Workaround

For graphics-intensive OpenCL applications, use a vGPU type with more frame buffer.

3.5. In pass through mode, all GPUs connected to each other through NVLink must be assigned to the same VM

Description

In pass through mode, all GPUs connected to each other through NVLink must be assigned to the same VM. If a subset of GPUs connected to each other through NVLink is passed through to a VM, unrecoverable error XID 74 occurs when the VM is booted. This error corrupts the NVLink state on the physical GPUs and, as a result, the NVLink bridge between the GPUs is unusable.

Workaround

Restore the NVLink state on the physical GPUs by resetting the GPUs or rebooting the hypervisor host.

3.6. vGPU profiles with 512 Mbytes or less of frame buffer support only 1 virtual display head on Windows 10

Description

To reduce the possibility of memory exhaustion, vGPU profiles with 512 Mbytes or less of frame buffer support only 1 virtual display head on a Windows 10 guest OS.

The following vGPU profiles have 512 Mbytes or less of frame buffer:

• Tesla M6-0B, M6-0Q
• Tesla M10-0B, M10-0Q
• Tesla M60-0B, M60-0Q

Workaround

Use a profile that supports more than 1 virtual display head and has at least 1 Gbyte of frame buffer.

3.7. NVENC requires at least 1 Gbyte of frame buffer

Description

Using the frame buffer for the NVIDIA hardware-based H.264/HEVC video encoder (NVENC) may cause memory exhaustion with vGPU profiles that have 512 Mbytes or less of frame buffer. To reduce the possibility of memory exhaustion, NVENC is disabled on profiles that have 512 Mbytes or less of frame buffer. Application GPU acceleration remains fully supported and available for all profiles, including profiles with 512 MBytes or less of frame buffer. NVENC support from both Citrix and VMware is a recent feature and, if you are using an older version, you should experience no change in functionality.

The following vGPU profiles have 512 Mbytes or less of frame buffer:

• Tesla M6-0B, M6-0Q
• Tesla M10-0B, M10-0Q
• Tesla M60-0B, M60-0Q

Workaround

If you require NVENC to be enabled, use a profile that has at least 1 Gbyte of frame buffer.

3.8. VM failures or crashes on servers with 1 TiB or more of system memory

Description

Support for vGPU and vSGA is limited to servers with less than 1 TiB of system memory. On servers with 1 TiB or more of system memory, VM failures or crashes may occur. For example, when Citrix Virtual Apps and Desktops is used with a Windows 7 guest OS, a blue screen crash may occur. However, support for vDGA is not affected by this limitation. Depending on the version of NVIDIA vGPU software that you are using, the log file on the VMware vSphere host might also report the following errors:

            

            
2016-10-27T04:36:21.128Z cpu74:70210)DMA: 1935: Unable to perform element mapping: DMA mapping could not be completed
2016-10-27T04:36:21.128Z cpu74:70210)Failed to DMA map address 0x118d296c000 (0x4000): Can't meet address mask of the device..
2016-10-27T04:36:21.128Z cpu74:70210)NVRM: VM: nv_alloc_contig_pages: failed to allocate memory
        

This limitation applies only to systems with supported GPUs based on the Maxwell architecture: Tesla M6, Tesla M10, and Tesla M60.

Resolution

Limit the amount of system memory on the server to 1 TiB minus 16 GiB.

1. Set memmapMaxRAMMB to 1032192, which is equal to 1048576 minus 16384.

   For detailed instructions, see Set Advanced Host Attributes in the VMware vSphere documentation.

2. Reboot the server.

If the problem persists, contact your server vendor for the recommended system memory configuration with NVIDIA GPUs.

3.9. VM running an incompatible NVIDIA vGPU guest driver fails to initialize vGPU when booted

Description

A VM running a version of the NVIDIA guest VM driver that is incompatible with the current release of Virtual GPU Manager will fail to initialize vGPU when booted on a VMware vSphere platform running that release of Virtual GPU Manager.

A guest VM driver is incompatible with the current release of Virtual GPU Manager in either of the following situations:

• The guest driver is from a release in a branch two or more major releases before the current release, for example release 9.4.

  In this situation, the VMware vSphere VM’s log file reports the following error:

  Copy

  Copied!

              
  
              
  vmiop_log: (0x0): Incompatible Guest/Host drivers: Guest VGX version is older than the minimum version supported by the Host. Disabling vGPU.
          

• The guest driver is from a later release than the Virtual GPU Manager.

  In this situation, the VMware vSphere VM’s log file reports the following error:

  Copy

  Copied!

              
  
              
  vmiop_log: (0x0): Incompatible Guest/Host drivers: Guest VGX version is newer than the maximum version supported by the Host. Disabling vGPU.
          

In either situation, the VM boots in standard VGA mode with reduced resolution and color depth. The NVIDIA virtual GPU is present in Windows Device Manager but displays a warning sign, and the following device status:

            

            
Windows has stopped this device because it has reported problems. (Code 43)
        

Resolution

Install a release of the NVIDIA guest VM driver that is compatible with current release of Virtual GPU Manager.

3.10. Single vGPU benchmark scores are lower than pass-through GPU

Description

A single vGPU configured on a physical GPU produces lower benchmark scores than the physical GPU run in pass-through mode.

Aside from performance differences that may be attributed to a vGPU’s smaller frame buffer size, vGPU incorporates a performance balancing feature known as Frame Rate Limiter (FRL). On vGPUs that use the best-effort scheduler, FRL is enabled. On vGPUs that use the fixed share or equal share scheduler, FRL is disabled.

FRL is used to ensure balanced performance across multiple vGPUs that are resident on the same physical GPU. The FRL setting is designed to give good interactive remote graphics experience but may reduce scores in benchmarks that depend on measuring frame rendering rates, as compared to the same benchmarks running on a pass-through GPU.

Resolution

FRL is controlled by an internal vGPU setting. On vGPUs that use the best-effort scheduler, NVIDIA does not validate vGPU with FRL disabled, but for validation of benchmark performance, FRL can be temporarily disabled by adding the configuration parameter pciPassthru0.cfg.frame_rate_limiter in the VM’s advanced configuration options.

1. Select Edit Settings.
2. In Edit Settings window, select the VM Options tab.
3. From the Advanced drop-down list, select Edit Configuration.
4. In the Configuration Parameters dialog box, click Add Row.
5. In the Name field, type the parameter name pciPassthru0.cfg.frame_rate_limiter, in the Value field type 0, and click OK.

   

With this setting in place, the VM’s vGPU will run without any frame rate limit. The FRL can be reverted back to its default setting by setting pciPassthru0.cfg.frame_rate_limiter to 1 or by removing the parameter from the advanced settings.

3.11. VMs configured with large memory fail to initialize vGPU when booted

Description

When starting multiple VMs configured with large amounts of RAM (typically more than 32GB per VM), a VM may fail to initialize vGPU. In this scenario, the VM boots in VMware SVGA mode and doesn’t load the NVIDIA driver. The NVIDIA vGPU software GPU is present in Windows Device Manager but displays a warning sign, and the following device status:

            

            
Windows has stopped this device because it has reported problems. (Code 43)
        

The VMware vSphere VM’s log file contains these error messages:

            

            
vthread10|E105: NVOS status 0x29
vthread10|E105: Assertion Failed at 0x7620fd4b:179
vthread10|E105: 8 frames returned by backtrace
 ...
vthread10|E105: VGPU message 12 failed, result code: 0x29
...
vthread10|E105: NVOS status 0x8
vthread10|E105: Assertion Failed at 0x7620c8df:280
vthread10|E105: 8 frames returned by backtrace
...
vthread10|E105: VGPU message 26 failed, result code: 0x8
        

Resolution

vGPU reserves a portion of the VM’s framebuffer for use in GPU mapping of VM system memory. The reservation is sufficient to support up to 32GB of system memory, and may be increased to accommodate up to 64GB by adding the configuration parameter pciPassthru0.cfg.enable_large_sys_mem in the VM’s advanced configuration options

1. Select Edit Settings.
2. In Edit Settings window, select the VM Options tab.
3. From the Advanced drop-down list, select Edit Configuration.
4. In the Configuration Parameters dialog box, click Add Row.
5. In the Name field, type the parameter name pciPassthru0.cfg.enable_large_sys_mem, in the Value field type 1, and click OK.

With this setting in place, less GPU framebuffer is available to applications running in the VM. To accommodate system memory larger than 64GB, the reservation can be further increased by adding pciPassthru0.cfg.extra_fb_reservation in the VM’s advanced configuration options, and setting its value to the desired reservation size in megabytes. The default value of 64M is sufficient to support 64 GB of RAM. We recommend adding 2 M of reservation for each additional 1 GB of system memory. For example, to support 96 GB of RAM, set pciPassthru0.cfg.extra_fb_reservation to 128.

The reservation can be reverted back to its default setting by setting pciPassthru0.cfg.enable_large_sys_mem to 0, or by removing the parameter from the advanced settings.

Only resolved issues that have been previously noted as known issues or had a noticeable user impact are listed. The summary and description for each resolved issue indicate the effect of the issue on NVIDIA vGPU software before the issue was resolved.

Issues Resolved in Release 12.4

Issues Resolved in Release 12.3

No resolved issues are reported in this release for VMware vSphere.

Issues Resolved in Release 12.2

Issues Resolved in Release 12.1

Issues Resolved in Release 12.0

No resolved issues are reported in this release for VMware vSphere.

5.1. VM fails after a second vGPU is assigned to it

Description

After a second vGPU is added to a VM and the VM is restarted, the VM fails. NVIDIA vGPU software supports up to a maximum of four vGPUs per VM on VMware vSphere Hypervisor (ESXi).

When this issue occurs, the following messages are written to the log file on the hypervisor host:

            

            
2021-09-27T17:11:42.303Z| vthread-2105551| | I005: vmiop_log: (0x0): Start restoring vGPU state ...
2021-09-27T17:11:43.465Z| vcpu-0| | E002: vmiop_log: (0x0): Deferred restore for RPCs cannot continue, since restore data was not saved
2021-09-27T17:11:43.465Z| vcpu-0| | E002: vmiop_log: (0x0): Deferred call for vmiopd_restore_rpc_data failed at un-stun!
2021-09-27T17:11:43.465Z| vcpu-0| | E002: vmiop_log: (0x0): Failed to complete restore for deferred functions.
2021-09-27T18:44:27.034Z| vthread-2105550| | E002: vmiop_log: (0x0): VGPU message 1 failed, guest VGX version is already initialized...
2021-09-27T18:44:27.034Z| vthread-2105550| | E002: vmiop_log: (0x0): VGPU message 1 failed, result code: 0x40
...
2021-09-27T18:44:35.359Z| vthread-2105550| | I005: vmiop_log: (0x0): Guest driver unloaded!
        

Workaround

To avoid this issue, create your VMs in EFI mode.

If you encounter this issue with a VM that was created in legacy BIOS mode, shut down and restart the VM or power off the VM and power it on again.

Status

Not an NVIDIA bug

Ref. #

3386681

5.2. 12.1-12.3 Only: vGPU Manager does not load on ESXi for the NVIDIA A100 HGX 80GB GPU

Description

The NVIDIA Virtual GPU Manager does not load on VMware vSphere ESXi for the NVIDIA A100 HGX 80GB GPU. This issue prevents NVIDIA A100 HGX 80GB GPU from being used for vGPU deployments on VMware vSphere ESXi.

Status

Resolved in NVIDIA vGPU software 12.4

Ref. #

200702059

5.3. Since 12.3: NVENC does not work with Teradici Cloud Access Software on Windows

Description

The NVIDIA hardware-based H.264/HEVC video encoder (NVENC) does not work with Teradici Cloud Access Software on Windows. This issue affects NVIDIA vGPU and GPU pass through deployments.

This issue occurs because the check that Teradici Cloud Access Software performs on the DLL signer name is case sensitive and NVIDIA recently changed the case of the company name in the signature certificate.

Status

Not an NVIDIA bug

This issue is resolved in the latest 21.07 and 21.03 Teradici Cloud Access Software releases.

Ref. #

200749065

5.4. When a licensed client deployed by using VMware instant clone technology is destroyed, it does not return the license

Description

When a user logs out of a VM deployed by using VMware Horizon instant clone technology, the VM is deleted and OS is not shut down cleanly. The NVIDIA vGPU software license that was being used by the VM is not returned to the license server, which could cause the license server to run out of licenses.

Workaround

Deploy the instant-clone desktop pool with the following options:

• Floating user assignment
• All Machines Up-Front provisioning

This configuration will allow the MAC address to be reused on the newly cloned VMs.

For more information, refer to the documentation for the version of VMware Horizon that you are using:

• VMware Horizon 8: Worksheet for Creating an Instant-Clone Desktop Pool in Horizon Console
• VMware Horizon 7: Worksheet for Creating an Instant-Clone Desktop Pool in Horizon Console

Status

Not an NVIDIA bug

Ref. #

200744338

5.5. A licensed client might fail to acquire a license if a proxy is set

Description

If a proxy is set with a system environment variable such as HTTP_PROXY or HTTPS_PROXY, a licensed client might fail to acquire a license.

Workaround

Perform this workaround on each affected licensed client.

1. Add the address of the NVIDIA vGPU software license server to the system environment variable NO_PROXY.

   The address must be specified exactly as it is specified in the client's license server settings either as a fully-qualified domain name or an IP address. If the NO_PROXY environment variable contains multiple entries, separate the entries with a comma (,).

   If high availability is configured for the license server, add the addresses of the primary license server and the secondary license server to the system environment variable NO_PROXY.

2. Restart the NVIDIA driver service that runs the core NVIDIA vGPU software logic.

   • On Windows, restart the NVIDIA Display Container service.
   • On Linux, restart the nvidia-gridd service.

Status

Closed

Ref. #

200704733

5.6. Session connection fails with four 4K displays and NVENC enabled on a 2Q, 3Q, or 4Q vGPU

Description

Desktop session connections fail for a 2Q, 3Q, or 4Q vGPU that is configured with four 4K displays and for which the NVIDIA hardware-based H.264/HEVC video encoder (NVENC) is enabled. This issue affects only Teradici Cloud Access Software sessions on Linux guest VMs.

This issue is accompanied by the following error message:

            

            
This Desktop has no resources available or it has timed out
        

This issue is caused by insufficient frame buffer.

Workaround

Ensure that sufficient frame buffer is available for all the virtual displays that are connected to a vGPU by changing the configuration in one of the following ways:

• Reducing the number of virtual displays. The number of 4K displays supported with NVENC enabled depends on the vGPU.

  vGPU	4K Displays Supported with NVENC Enabled
  2Q	1
  3Q	2
  4Q	3

• Disabling NVENC. The number of 4K displays supported with NVENC disabled depends on the vGPU.

  vGPU	4K Displays Supported with NVENC Disabled
  2Q	2
  3Q	2
  4Q	4

• Using a vGPU type with more frame buffer. Four 4K displays with NVENC enabled on any Q-series vGPU with at least 6144 MB of frame buffer are supported.

Status

Not an NVIDIA bug

Ref. #

200701959

5.7. Disconnected sessions cannot be reconnected or might be reconnected very slowly with NVWMI installed

Description

Disconnected sessions cannot be reconnected or might be reconnected very slowly when the NVIDIA Enterprise Management Toolkit (NVWMI) is installed. This issue affects Citrix Virtual Apps and Desktops and VMware Horizon sessions on Windows guest VMs.

Workaround

Uninstall NVWMI.

Status

Open

Ref. #

3262923

5.8. Windows VM crashes during Custom (Advanced) driver upgrade

Description

When the NVIDIA vGPU software graphics driver in a Windows VM is upgraded with the Custom (Advanced) option selected, the VM crashes.

Status

Open

Ref. #

200700291

5.9. Since 12.1: VMs with vGPUs on GPUs based on the NVIDIA Ampere architecture fail to power on

Description

An otherwise correctly configured VMware vSphere ESXi 7.0 Update 2 server fails to boot VMs with vGPUs on GPUs based on the NVIDIA Ampere if the server being managed by a version of VMware vCenter Server older than 7.0.2. This version of VMware vCenter is released with ESXi 7.0 VMware vSphere Update 2.

When this issue occurs, the following error message is seen:

            

            
Insufficient resources. One or more devices (pciPassthru0) required by VM vm-name are not available on host host-name
        

Workaround

Use VMware vCenter Server 7.0.2 or a later compatible update

Status

Open

5.10. 12.1 Only: Purple screen crash occurs after the Xorg service is started in a VM with multiple vGPUs

Description

After the Xorg service is started in a guest VM to which multiple vGPUs are assigned, a purple screen crash occurs on the hypervisor host. This issue affects only the NVIDIA A100 HGX 80GB, NVIDIA A100 PCIe 40GB, and NVIDIA A100 HGX 40GB GPUs.

When this issue occurs, the following error message is written to the vmkernel.log file:

            

            
IO Protection for [0x10bc5c000, 0x10bc5d000) is not READWRITE
        

Status

Resolved in NVIDIA vGPU software 12.2

Ref. #

200694284

5.11. Migrating a VM with a Tesla T4 vGPU between a host running NVIDIA vGPU software 11.3 and a host running a different release fails

Description

Migrating a VM with a Tesla T4 vGPU between hosts where one host is running NVIDIA vGPU software 11.3 and the other host is running a different release fails. After the migration, the destination host and VM become unstable.

When this issue occurs, error message XID error 38 is written to the VMware vSphere VM’s log file vmware.log in the guest VM’s storage directory. Depending on the host configurations, the following messages might also be written to the log file:

• XID error 43
• VGPU message 58
• VGPU message 4

Version

This issue affects migrations between a host that is running NVIDIA vGPU software release 11.3 and a host that is running a different release. The issue affects migrations to and from the host that is running NVIDIA vGPU software release 11.3.

This issue does not affect migrations between two hosts that are both running NVIDIA vGPU software 11.3.

Workaround

Avoid migrating a VM between a host that is running NVIDIA vGPU software 11.3 and a host that is running a different NVIDIA vGPU software release.

Status

Closed

Ref. #

200691763

200735219

5.12. 12.1 Only: Migration hangs with NVIDIA A40 and RTX A6000 GPUs

Description

vGPU migration hangs with NVIDIA A40 and RTX A6000 GPUs.

Status

Resolved in NVIDIA vGPU software 12.2

Ref. #

3207447

5.13. 12.0 Only: Sessions freeze when Adobe Premiere with the Adobe Mercury Engine is used

Description

Sessions freeze when Adobe Premiere with the Adobe Mercury Engine is used. The session can freeze after 15-20 minutes of use or when a project is being exported.

Workaround

Status

Resolved in NVIDIA vGPU software 12.1

Ref. #

3230997

5.14. 12.0 Only: Issues occur when Blackmagic Design DaVinci Resolve is used

Description

Multiple issues, such as application crashes, application instability, and session freezes, occur when Blackmagic Design DaVinci Resolve is used. CUDA error 702 might also be observed.

Status

Resolved in NVIDIA vGPU software 12.1

Ref. #

3225521

5.15. 12.0, 12.1 Only: NVIDIA vGPU software graphics driver installation fails in Ubuntu guest VMs

Description

Installation of the NVIDIA vGPU software graphics driver from a .run file fails in Ubuntu guest VMs that are running Linux kernel version 5.8 or later.

Version

Ubuntu 20.04 LTS and 18.04 with Linux kernel version 5.8 or later

Workaround

Revert to a Linux kernel version earlier than 5.8.

Status

Resolved in NVIDIA vGPU software 12.2

Ref. #

3226853

5.16. NVML fails to initialize with unknown error

Description

After NVIDIA Virtual GPU manager is installed on VMware vSphere Hypervisor (ESXi) 6.7, the nvidia-smi command fails with the error Failed to initialize NVML: Unknown Error.

Version

VMware vSphere Hypervisor (ESXi) 6.7

Workaround

Apply VMware ESXi 6.7, Patch Release ESXi670-202011002, build 17167734 or later from VMware.

Status

Not an NVIDIA bug

Ref. #

3237970

5.17. VM hangs after vGPU migration to a host running a newer vGPU manager version

Description

When a VM configured with a Tesla V100 or Tesla T4 vGPU is migrated from a host that is running a vGPU manager 11 release to a host that is running vGPU manager 12.0, the VM hangs. When this issue occurs, XID error 31 is written to the log files on the destination hypervisor host. After the migration, the destination host and VM become unstable.

Version

This issue affects migration from a host that is running a vGPU manager 11 release, such as 11.3 or 11.2, to a host that is running vGPU manager 12.0.

Status

Open

Ref. #

200691445

5.18. Idle Teradici Cloud Access Software session disconnects from Linux VM

Description

After a Teradici Cloud Access Software session has been idle for a short period of time, the session disconnects from the VM. When this issue occurs, the error messages NVOS status 0x19 and vGPU Message 21 failed are written to the log files on the hypervisor host. This issue affects only Linux guest VMs.

Status

Open

Ref. #

200689126

5.19. GPU Operator doesn't support vGPU on GPUs based on architectures before NVIDIA Turing

Description

NVIDIA GPU Operator doesn't support vGPU deployments on GPUs based on architectures before the NVIDIA Turing™ architecture. This issue is caused by the omission of version information for the vGPU manager from the configuration information that GPU Operator requires. Without this information, GPU Operator does not deploy the NVIDIA driver container because the container cannot determine if the driver is compatible with the vGPU manager.

Status

Open

Ref. #

3227576

5.20. vGPU migration to a host running an older vGPU manager version fails with multiple errors

Description

When a VM configured with a Tesla V100 or Tesla T4 vGPU is migrated from a host that is running vGPU manager 12.0 to a host that is running a vGPU manager 11 release, the migration fails. When this issue occurs, XID errors 44 and 43, timeout detection and recovery (TDR) errors, NVOS status 0x65, and vGPU message 22 are written to the log files on the destination hypervisor host. After the migration, the destination host and VM become unstable.

Version

This issue affects migration from a host that is running vGPU manager 12.0 to a host that is running a vGPU manager 11 release, such as 11.3 or 11.2.

This issue affects only VMs configured with a vGPU that resides on a Tesla V100 GPU or Tesla T4 GPU.

Status

Open

Ref. #

200690020

5.21. Idle NVIDIA A100, NVIDIA A40, and NVIDIA A10 GPUs show 100% GPU utilization

Description

The nvidia-smi command shows 100% GPU utilization for NVIDIA A100, NVIDIA A40, and NVIDIA A10 GPUs even if no vGPUs have been configured or no VMs are running.

            

            
[root@host ~]# nvidia-smi
Fri Oct 29 11:45:28 2021
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 460.107   Driver Version: 460.107  CUDA Version: 11.2       |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  A100-PCIE-40GB      On   | 00000000:5E:00.0 Off |                    0 |
| N/A   50C    P0    97W / 250W |      0MiB / 40537MiB |    100%      Default |
|                               |                      |             Disabled |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+
        

Workaround

Boot any VMs that are configured with a vGPU that resides on the GPU.

After this workaround has been completed, the nvidia-smi command shows 0% GPU utilization for affected GPUs when they are idle.

            

            
root@host ~]# nvidia-smi
Fri Oct 29 11:47:38 2021
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 460.107   Driver Version: 460.107  CUDA Version: 11.2       |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  A100-PCIE-40GB      On   | 00000000:5E:00.0 Off |                    0 |
| N/A   50C    P0    97W / 250W |      0MiB / 40537MiB |      0%      Default |
|                               |                      |             Disabled |
+-------------------------------+----------------------+----------------------+

+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+
        

Status

Open

Ref. #

200605527

5.22. Driver upgrade in a Linux guest VM with multiple vGPUs might fail

Description

Upgrading the NVIDIA vGPU software graphics driver in a Linux guest VM with multiple vGPUs might fail. This issue occurs if the driver is upgraded by overinstalling the new release of the driver on the current release of the driver while the nvidia-gridd service is running in the VM.

Workaround

1. Stop the nvidia-gridd service.
2. Try again to upgrade the driver.

Status

Open

Ref. #

200633548

5.23. NVIDIA Control Panel fails to start if launched too soon from a VM without licensing information

Description

If NVIDIA licensing information is not configured on the system, any attempt to start NVIDIA Control Panel by right-clicking on the desktop within 30 seconds of the VM being started fails.

Workaround

Restart the VM and wait at least 30 seconds before trying to launch NVIDIA Control Panel.

Status

Open

Ref. #

200623179

5.24.  Citrix Virtual Apps and Desktops session corruption occurs in the form of residual window borders

Description

When a window is dragged across the desktop in a Citrix Virtual Apps and Desktops session, corruption of the session in the form of residual window borders occurs.

Version

This issue affects only Citrix Virtual Apps and Desktops version 7 2003

Workaround

Use Citrix Virtual Apps and Desktops version 7 1912 or 2006.

Status

Not an NVIDIA bug

Ref. #

200608675

5.25. VMware Horizon clients cannot connect to a Windows 10 2004 VM with multiple displays

Description

Some VMware Horizon clients cannot connect to a Windows 10 2004 VM with multiple displays. When this issue occurs, the VM becomes unusable and clients cannot connect to the VM even if only a single display is connected to it.

This issue occurs because the desktop capture mechanism for the affected VMware Horizon clients is provided by NVIDIA^® Frame Buffer Capture (NVFBC) and NVFBC is deprecated on Windows 10 starting with Windows 10 October 2019 Update. For more information, see NVFBC Windows 10 Support Deprecation Technical Bulletin (PDF).

Version

This issue affects only Windows 10 May 2020 Update (2004) guest VMs.

Workaround

Contact VMware to obtain a version of VMware Horizon for which the desktop capture mechanism is not provided by NVFBC.

Status

Not an NVIDIA bug

Ref. #

200607827

5.26. Suspend and resume between hosts running different versions of the vGPU manager fails

Description

Suspending a VM configured with vGPU on a host running one version of the vGPU manager and resuming the VM on a host running a version from an older main release branch fails. For example, suspending a VM on a host that is running the vGPU manager from release 12.4 and resuming the VM on a host running the vGPU manager from release 11.6 fails. When this issue occurs, the error One or more devices (pciPassthru0) required by VM vm-name are not available on host host-name is reported on VMware vCenter Server.

Status

Not an NVIDIA bug

Ref. #

200602087

5.27. On Linux, a VMware Horizon 7.12 session freezes after a switch to full screen

Description

On a Linux VM configured with a -1Q vGPU, one 4K display, and VMware Horizon 7.12, the VMware Horizon session might become unresponsive after a switch from large screen (windowed) to full screen. When this issue occurs, the VMware vSphere VM’s log file contains the error message Unable to set requested topology.

Version

This issue affects deployments that use VMware Horizon 7.12.

Workaround

Use VMware Horizon 7.11.

Status

Open

Ref. #

200617112

5.28. On Linux, a VMware Horizon 7.12 session with two 4K displays freezes

Description

On a Linux VM configured with a -1Q vGPU, two 4K displays, and VMware Horizon 7.12, the VMware Horizon session might become unresponsive. When this issue occurs, the VMware vSphere VM’s log file contains the error message Failed to setup capture session (error 8). Unable to allocate video memory.

Version

This issue affects deployments that use VMware Horizon 7.12.

Workaround

Use VMware Horizon 7.11 or a vGPU with more frame buffer.

Status

Open

Ref. #

200617081

5.29. On Linux, the frame rate might drop to 1 after several minutes

Description

On Linux, the frame rate might drop to 1 frame per second (FPS) after NVIDIA vGPU software has been running for several minutes. Only some applications are affected, for example, glxgears. Other applications, such as Unigine Heaven, are not affected. This behavior occurs because Display Power Management Signaling (DPMS) for the Xorg server is enabled by default and the display is detected to be inactive even when the application is running. When DPMS is enabled, it enables power saving behavior of the display after several minutes of inactivity by setting the frame rate to 1 FPS.

Workaround

1. If necessary, stop the Xorg server.

   Copy

   Copied!

               
   
               
   # /etc/init.d/xorg stop
           

2. In a plain text editor, edit the /etc/X11/xorg.conf file to set the options to disable DPMS and disable the screen saver.

   1. In the Monitor section, set the DPMS option to false.
      Copy

      Copied!

                  
      
                  
      Option "DPMS" "false"
              

   2. At the end of the file, add a ServerFlags section that contains option to disable the screen saver.
      Copy

      Copied!

                  
      
                  
      Section "ServerFlags"
          Option "BlankTime" "0"
        EndSection
              

   3. Save your changes to /etc/X11/xorg.conf file and quit the editor.

3. Start the Xorg server.

   Copy

   Copied!

               
   
               
   # etc/init.d/xorg start
           

Status

Open

Ref. #

200605900

5.30. Frame buffer consumption grows with VMware Horizon over Blast Extreme

Description

When VMware Horizon is used with the Blast Extreme display protocol, frame buffer consumption increases over time after multiple disconnections from and reconnections to a VM. This issue occurs even if the VM is in an idle state and no graphics applications are running.

Workaround

Reboot the VM.

Status

Not an NVIDIA bug

Ref. #

200602520

5.31. DWM crashes randomly occur in Windows VMs

Description

Desktop Windows Manager (DWM) crashes randomly occur in Windows VMs, causing a blue-screen crash and the bug check CRITICAL_PROCESS_DIED. Computer Management shows problems with the primary display device.

Version

This issue affects Windows 10 1809, 1903 and 1909 VMs.

Status

Not an NVIDIA bug

Ref. #

2730037

5.32. Remote desktop session freezes with assertion failure and XID error 43 after migration

Description

After multiple VMs configured with vGPU on a single hypervisor host are migrated simultaneously, the remote desktop session freezes with an assertion failure and XID error 43. This issue affects only GPUs that are based on the Volta GPU architecture. It does not occur if only a single VM is migrated.

When this error occurs, the following error messages are logged to the VMware vSphere Hypervisor (ESXi) log file:

            

            
Jan  3 14:35:48 ch81-m1 vgpu-12[8050]: error: vmiop_log: NVOS status 0x1f
Jan  3 14:35:48 ch81-m1 vgpu-12[8050]: error: vmiop_log: Assertion Failed at 0x4b8cacf6:286
...
Jan  3 14:35:59 ch81-m1 vgpu-12[8050]: error: vmiop_log: (0x0): XID 43 detected on physical_chid:0x174, guest_chid:0x14
        

Status

Open

Ref. #

200581703

5.33. Citrix Virtual Apps and Desktops session freezes when the desktop is unlocked

Description

When a Citrix Virtual Apps and Desktops session that is locked is unlocked by pressing Ctrl+Alt+Del, the session freezes. This issue affects only VMs that are running Microsoft Windows 10 1809 as a guest OS.

Version

Microsoft Windows 10 1809 guest OS

Workaround

Restart the VM.

Status

Not an NVIDIA bug

Ref. #

2767012

5.34. NVIDIA vGPU software graphics driver fails after Linux kernel upgrade with DKMS enabled

Description

After the Linux kernel is upgraded (for example by running sudo apt full-upgrade) with Dynamic Kernel Module Support (DKMS) enabled, the nvidia-smi command fails to run. If DKMS is enabled, an upgrade to the Linux kernel triggers a rebuild of the NVIDIA vGPU software graphics driver. The rebuild of the driver fails because the compiler version is incorrect. Any attempt to reinstall the driver fails because the kernel fails to build.

When the failure occurs, the following messages are displayed:

            

            
-> Installing DKMS kernel module:
        ERROR: Failed to run `/usr/sbin/dkms build -m nvidia -v  460.32.03 -k 5.3.0-28-generic`: 
        Kernel preparation unnecessary for this kernel. Skipping...
        Building module:
        cleaning build area...
        'make' -j8 NV_EXCLUDE_BUILD_MODULES='' KERNEL_UNAME=5.3.0-28-generic IGNORE_CC_MISMATCH='' modules...(bad exit status: 2)
        ERROR (dkms apport): binary package for nvidia:  460.32.03 not found
        Error! Bad return status for module build on kernel: 5.3.0-28-generic (x86_64)
        Consult /var/lib/dkms/nvidia/ 460.32.03/build/make.log for more information.
        -> error.
        ERROR: Failed to install the kernel module through DKMS. No kernel module was installed;
        please try installing again without DKMS, or check the DKMS logs for more information.
        ERROR: Installation has failed. Please see the file '/var/log/nvidia-installer.log' for details.
        You may find suggestions on fixing installation problems in the README available on the Linux driver download page at www.nvidia.com.
        

Workaround

When installing the NVIDIA vGPU software graphics driver with DKMS enabled, use one of the following workarounds:

• Before running the driver installer, install the dkms package, then run the driver installer with the -dkms option.
• Run the driver installer with the --no-cc-version-check option.

Status

Not a bug.

Ref. #

2836271

5.35. Red Hat Enterprise Linux and CentOS 6 VMs hang during driver installation

Description

During installation of the NVIDIA vGPU software graphics driver in a Red Hat Enterprise Linux or CentOS 6 guest VM, a kernel panic occurs, and the VM hangs and cannot be rebooted. This issue is observed on older Linux kernels when the NVIDIA device is using message-signaled interrupts (MSIs).

Version

This issue affects the following guest OS releases:

• Red Hat Enterprise Linux 6.6 and later compatible 6.x versions
• CentOS 6.6 and later compatible 6.x versions

Workaround

1. Disable MSI in the guest VM to fall back to INTx interrupts by adding the following line to the file /etc/modprobe.d/nvidia.conf:

   Copy

   Copied!

               
   
               
   options nvidia NVreg_EnableMSI=0
           

   If the file /etc/modprobe.d/nvidia.conf does not exist, create it.

2. Install the NVIDIA vGPU Software graphics driver in the guest VM.

Status

Closed

Ref. #

200556896

5.36. Publisher not verified warning during Windows 7 driver installation

Description

During installation of the NVIDIA vGPU software graphics driver for Windows on Windows 7, Windows warns that it can't verify the publisher of the driver software. If Device Manager is used to install the driver, Device Manager warns that the driver is not digitally signed. If you install the driver, error 52 (CM_PROB_UNSIGNED_DRIVER) occurs.

This issue occurs because Microsoft is no longer dual signing WHQL-tested software binary files by using the SHA-1 and SHA-2 hash algorithms. Instead, WHQL-tested software binary files are signed only by using the SHA-2 hash algorithm. All NVIDIA vGPU software graphics drivers for Windows are WHQL tested.

By default, Windows 7 systems cannot recognize signatures that were created by using the SHA-2 hash algorithm. As a result, software binary files that are signed only by using the SHA-2 hash algorithm are considered unsigned.

For more information, see 2019 SHA-2 Code Signing Support requirement for Windows and WSUS on the Microsoft Windows support website.

Version

Windows 7

Workaround

If you experience this issue, install the following updates and restart the VM or host before installing the driver:

• Servicing stack update (SSU) (KB4490628)
• SHA-2 update (KB4474419)

Status

Not a bug

5.37. Tesla T4 is enumerated as 32 separate GPUs by VMware vSphere ESXi

Description

Some servers, for example, the Dell R740, do not configure SR-IOV capability if the SR-IOV SBIOS setting is disabled on the server. If the SR-IOV SBIOS setting is disabled on such a server that is being used with the Tesla T4 GPU, VMware vSphere ESXi enumerates the Tesla T4 as 32 separate GPUs. In this state, you cannot use the GPU to configure a VM with NVIDIA vGPU or for GPU pass through.

Workaround

Ensure that the SR-IOV SBIOS setting is enabled on the server.

Status

Not an NVIDIA bug

A fix is available from VMware in VMware vSphere ESXi 7.0 Update 2.

Ref. #

2697051

5.38. VMware vCenter shows GPUs with no available GPU memory

Description

VMware vCenter shows some physical GPUs as having 0.0 B of available GPU memory. VMs that have been assigned vGPUs on the affected physical GPUs cannot be booted. The nvidia-smi command shows the same physical GPUs as having some GPU memory available.

Workaround

Stop and restart the Xorg service and nv-hostengine on the ESXi host.

1. Stop all running VM instances on the host.

2. Stop the Xorg service.

   Copy

   Copied!

               
   
               
   [root@esxi:~] /etc/init.d/xorg stop
           

3. Stop nv-hostengine.

   Copy

   Copied!

               
   
               
   [root@esxi:~] nv-hostengine -t
           

4. Wait for 1 second to allow nv-hostengine to stop.

5. Start nv-hostengine.

   Copy

   Copied!

               
   
               
   [root@esxi:~] nv-hostengine -d
           

6. Start the Xorg service.

   Copy

   Copied!

               
   
               
   [root@esxi:~] /etc/init.d/xorg start
           

Status

Not an NVIDIA bug

A fix is available from VMware in VMware vSphere ESXi 6.7 U3. For information about the availability of fixes for other releases of VMware vSphere ESXi, contact VMware.

Ref. #

2644794

5.39. RAPIDS cuDF merge fails on NVIDIA vGPU

Description

The merge function of the RAPIDS cuDF GPU data frame library fails on NVIDIA vGPU. This function fails because RAPIDS uses the Unified Memory feature of CUDA, which NVIDIA vGPU does not support.

Status

Open

Ref. #

2642134

5.40. Users' sessions may freeze during vMotion migration of VMs configured with vGPU

Description

When vMotion is used to migrate a VM configured with vGPU to another host, users' sessions may freeze for up to several seconds during the migration.

These factors may increase the length of time for which a session freezes:

• Continuous use of the frame buffer by the workload, which typically occurs with workloads such as video streaming
• A large amount of vGPU frame buffer
• A large amount of system memory
• Limited network bandwidth

Workaround

Administrators can mitigate the effects on end users by avoiding migration of VMs configured with vGPU during business hours or warning end users that migration is about to start and that they may experience session freezes.

End users experiencing this issue must wait for their sessions to resume when the migration is complete.

Status

Open

Ref. #

2569578

5.41. Migration of VMs configured with vGPU stops before the migration is complete

Description

When a VM configured with vGPU is migrated to another host, the migration stops before it is complete. After the migration stops, the VM is no longer accessible.

This issue occurs if the ECC memory configuration (enabled or disabled) on the source and destination hosts are different. The ECC memory configuration on both the source and destination hosts must be identical.

Workaround

Reboot the hypervisor host to recover the VM. Before attempting to migrate the VM again, ensure that the ECC memory configuration on both the source and destination hosts are identical.

Status

Not an NVIDIA bug

A fix is that prevents the VM from becoming inaccessible is available from VMware in VMware vSphere Hypervisor (ESXi) 6.7 Update 3 patch 16075168-04282020. Even with this patch, migration of a VM configured with vGPU requires the ECC memory configuration on both the source and destination hosts to be identical.

Ref. #

200520027

5.42. ECC memory settings for a vGPU cannot be changed by using NVIDIA X Server Settings

Description

The ECC memory settings for a vGPU cannot be changed from a Linux guest VM by using NVIDIA X Server Settings. After the ECC memory state has been changed on the ECC Settings page and the VM has been rebooted, the ECC memory state remains unchanged.

Workaround

Use the nvidia-smi command in the guest VM to enable or disable ECC memory for the vGPU as explained in Virtual GPU Software User Guide.

If the ECC memory state remains unchanged even after you use the nvidia-smi command to change it, use the workaround in Changes to ECC memory settings for a Linux vGPU VM by nvidia-smi might be ignored.

Status

Open

Ref. #

200523086

5.43. Changes to ECC memory settings for a Linux vGPU VM by nvidia-smi might be ignored

Description

After the ECC memory state for a Linux vGPU VM has been changed by using the nvidia-smi command and the VM has been rebooted, the ECC memory state might remain unchanged.

This issue occurs when multiple NVIDIA configuration files in the system cause the kernel module option for setting the ECC memory state RMGuestECCState in /etc/modprobe.d/nvidia.conf to be ignored.

When the nvidia-smi command is used to enable ECC memory, the file /etc/modprobe.d/nvidia.conf is created or updated to set the kernel module option RMGuestECCState. Another configuration file in /etc/modprobe.d/ that contains the keyword NVreg_RegistryDwordsPerDevice might cause the kernel module option RMGuestECCState to be ignored.

Workaround

This workaround requires administrator privileges.

1. Move the entry containing the keyword NVreg_RegistryDwordsPerDevice from the other configuration file to /etc/modprobe.d/nvidia.conf.
2. Reboot the VM.

Status

Open

Ref. #

200505777

5.44. Black screens observed when a VMware Horizon session is connected to four displays

Description

When a VMware Horizon session with Windows 7 is connected to four displays, a black screen is observed on one or more displays.

This issue occurs because a VMware Horizon session does not support connections to four 4K displays with Windows 7.

Status

Not an NVIDIA bug

Ref. #

200503538

5.45. Quadro RTX 8000 and Quadro RTX 6000 GPUs can't be used with VMware vSphere ESXi 6.5

Description

Quadro RTX 8000 and Quadro RTX 6000 GPUs can't be used with VMware vSphere ESXi 6.5. If you attempt to use the Quadro RTX 8000 or Quadro RTX 6000 GPU with VMware vSphere ESXi 6.5, a purple-screen crash occurs after you install the NVIDIA Virtual GPU Manager.

Version

VMware vSphere ESXi 6.5

Workaround

Upgrade VMware vSphere ESXi to patch level ESXi 6.5 P04 (ESXi650-201912002, build 15256549) or later.

VMware resolved this issue in this patch for VMware vSphere ESXi.

Status

Not an NVIDIA bug

Ref. #

200491080

5.46. Vulkan applications crash in Windows 7 guest VMs configured with NVIDIA vGPU

Description

In Windows 7 guest VMs configured with NVIDIA vGPU, applications developed with Vulkan APIs crash or throw errors when they are launched. Vulkan APIs require sparse texture support, but in Windows 7 guest VMs configured with NVIDIA vGPU, sparse textures are not enabled.

In Windows 10 guest VMs configured with NVIDIA vGPU, sparse textures are enabled and applications developed with Vulkan APIs run correctly in these VMs.

Status

Open

Ref. #

200381348

5.47. Host core CPU utilization is higher than expected for moderate workloads

Description

When GPU performance is being monitored, host core CPU utilization is higher than expected for moderate workloads. For example, host CPU utilization when only a small number of VMs are running is as high as when several times as many VMs are running.

Workaround

Disable monitoring of the following GPU performance statistics:

• vGPU engine usage by applications across multiple vGPUs
• Encoder session statistics
• Frame buffer capture (FBC) session statistics
• Statistics gathered by performance counters in guest VMs

Status

Open

Ref. #

2414897

5.48. H.264 encoder falls back to software encoding on 1Q vGPUs with a 4K display

Description

On 1Q vGPUs with a 4K display, a shortage of frame buffer causes the H.264 encoder to fall back to software encoding.

Workaround

Use a 2Q or larger virtual GPU type to provide more frame buffer for each vGPU.

Status

Open

Ref. #

2422580

5.49. H.264 encoder falls back to software encoding on 2Q vGPUs with 3 or more 4K displays

Description

On 2Q vGPUs with three or more 4K displays, a shortage of frame buffer causes the H.264 encoder to fall back to software encoding.

This issue affects only vGPUs assigned to VMs that are running a Linux guest OS.

Workaround

Use a 4Q or larger virtual GPU type to provide more frame buffer for each vGPU.

Status

Open

Ref. #

200457177

5.50. Frame capture while the interactive logon message is displayed returns blank screen

Description

Because of a known limitation with NvFBC, a frame capture while the interactive logon message is displayed returns a blank screen.

An NvFBC session can capture screen updates that occur after the session is created. Before the logon message appears, there is no screen update after the message is shown and, therefore, a black screen is returned instead. If the NvFBC session is created after this update has occurred, NvFBC cannot get a frame to capture.

Workaround

Press Enter or wait for the screen to update for NvFBC to capture the frame.

Status

Not a bug

Ref. #

2115733

5.51. RDS sessions do not use the GPU with some Microsoft Windows Server releases

Description

When some releases of Windows Server are used as a guest OS, Remote Desktop Services (RDS) sessions do not use the GPU. With these releases, the RDS sessions by default use the Microsoft Basic Render Driver instead of the GPU. This default setting enables 2D DirectX applications such as Microsoft Office to use software rendering, which can be more efficient than using the GPU for rendering. However, as a result, 3D applications that use DirectX are prevented from using the GPU.

Version

• Windows Server 2019
• Windows Server 2016
• Windows Server 2012

Solution

Change the local computer policy to use the hardware graphics adapter for all RDS sessions.

1. Choose Local Computer Policy > Computer Configuration > Administrative Templates > Windows Components > Remote Desktop Services > Remote Desktop Session Host > Remote Session Environment.

2. Set the Use the hardware default graphics adapter for all Remote Desktop Services sessions option.

5.52. VMware vMotion fails gracefully under heavy load

Description

Migrating a VM configured with vGPU fails gracefully if the VM is running an intensive workload.

The error stack in the task details on the vSphere web client contains the following error message:

            

            
The migration has exceeded the maximum switchover time of 100 second(s). 
ESX has preemptively failed the migration to allow the VM to continue running on the source. 
To avoid this failure, either increase the maximum allowable switchover time or wait until
the VM is performing a less intensive workload.
        

Workaround

Increase the maximum switchover time by increasing the vmotion.maxSwitchoverSeconds option from the default value of 100 seconds.

For more information, see VMware Knowledge Base Article: vMotion or Storage vMotion of a VM fails with the error: The migration has exceeded the maximum switchover time of 100 second(s) (2141355).

Status

Not an NVIDIA bug

Ref. #

200416700

5.53. View session freezes intermittently after a Linux VM acquires a license

Description

In a Linux VM, the view session can sometimes freeze after the VM acquires a license.

Workaround

Resize the view session.

Status

Not an NVIDIA bug

Ref. #

200426961

5.54. When the scheduling policy is fixed share, GPU utilization is reported as higher than expected

Description

When the scheduling policy is fixed share, GPU engine utilization can be reported as higher than expected for a vGPU.

For example, GPU engine usage for six P40-4Q vGPUs on a Tesla P40 GPU might be reported as follows:

            

            
[root@localhost:~] nvidia-smi vgpu
Mon Aug 20 10:33:18 2018
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 390.42                 Driver Version: 390.42                    |
|-------------------------------+--------------------------------+------------+
| GPU  Name                     | Bus-Id                         | GPU-Util   |
|      vGPU ID    Name          | VM ID    VM Name               | vGPU-Util  |
|===============================+================================+============|
|   0  Tesla P40                | 00000000:81:00.0               |  99%       |
|      85109 GRID P40-4Q | 85110 win7-xmpl-146048-1 | 32%    |
|      87195 GRID P40-4Q | 87196 win7-xmpl-146048-2 | 39%    |
|      88095 GRID P40-4Q | 88096 win7-xmpl-146048-3 | 26%    |
|      89170      GRID P40-4Q   | 89171    win7-xmpl-146048-4    |      0%    |
|      90475      GRID P40-4Q   | 90476    win7-xmpl-146048-5    |      0%    |
|      93363      GRID P40-4Q   | 93364    win7-xmpl-146048-6    |      0%    |
+-------------------------------+--------------------------------+------------+
|   1  Tesla P40                | 00000000:85:00.0               |   0%       |
+-------------------------------+--------------------------------+------------+
        

The vGPU utilization of vGPU 85109 is reported as 32%. For vGPU 87195, vGPU utilization is reported as 39%. And for 88095, it is reported as 26%. However, the expected vGPU utilization of any vGPU should not exceed approximately 16.7%.

This behavior is a result of the mechanism that is used to measure GPU engine utilization.

Status

Open

Ref. #

2227591

5.55. nvidia-smi reports that vGPU migration is supported on all hypervisors

Description

The command nvidia-smi vgpu -m shows that vGPU migration is supported on all hypervisors, even hypervisors or hypervisor versions that do not support vGPU migration.

Status

Closed

Ref. #

200407230

5.56.  GPU resources not available error during VMware instant clone provisioning

Description

A GPU resources not available error might occur during VMware instant clone provisioning. On Windows VMs, a Video TDR failure - NVLDDMKM.sys error causes a blue screen crash.

This error occurs when options for VMware Virtual Shared Graphics Acceleration (vSGA) are set for a VM that is configured with NVIDIA vGPU. VMware vSGA is a feature of VMware vSphere that enables multiple virtual machines to share the physical GPUs on ESXi hosts and can be used as an alternative to NVIDIA vGPU.

Depending on the combination of options set, one of the following error messages is seen when the VM is powered on:

• Module ‘MKS’ power on failed.

  This message is seen when the following options are set:

  • Enable 3D support is selected.
  • 3D Renderer is set to Hardware
  • The graphics type of all GPUs on the ESXi host is Shared Direct.
• Hardware GPU resources are not available. The virtual machine will use software rendering.

  This message is seen when the following options are set:

  • Enable 3D support is selected.
  • 3D Renderer is set to Automatic.
  • The graphics type of all GPUs on the ESXi host is Shared Direct.

Resolution

If you want to use NVIDIA vGPU, unset any options for VMware vSGA that are set for the VM.

1. Ensure that the VM is powered off.
2. Open the vCenter Web UI.
3. In the vCenter Web UI, right-click the VM and choose Edit Settings.
4. Click the Virtual Hardware tab.
5. In the device list, expand the Video card node and de-select the Enable 3D support option.
6. Start the VM.

Status

Not a bug

Ref. #

2369683

5.57. VMs with 32 GB or more of RAM fail to boot with GPUs requiring 64 GB or more of MMIO space

Description

VMs with 32 GB or more of RAM fail to boot with GPUs that require 64 GB or more of MMIO space. VMs boot successfully with RAM allocations of less than 32 GB.

The following table lists the GPUs that require 64 GB or more of MMIO space and the amount of MMIO space that each GPU requires.

Version

This issue affects the following versions of VMware vSphere ESXi:

• 6.0 Update 3 and later updates
• 6.5 and later updates

Workaround

If you want to use a VM with 32 GB or more of RAM with GPUs that require 64 GB or more of MMIO space, use this workaround:

1. Create a VM to which less than 32 GB of RAM is allocated.

2. Choose VM Options > Advanced and set pciPassthru.use64bitMMIO="TRUE".

3. Allocate the required amount of RAM to the VM.

For more information, see VMware Knowledge Base Article: VMware vSphere VMDirectPath I/O: Requirements for Platforms and Devices (2142307).

Status

Not an NVIDIA bug

Resolved in VMware vSphere ESXi 6.7

Ref. #

2043171

5.58. Module load failed during VIB downgrade from R390 to R384

Description

Some registry keys are available only with the R390 Virtual GPU Manager, for example, NVreg_IgnoreMMIOCheck. If any keys that are available only with the R390 Virtual GPU Manager are set, the NVIDIA module fails to load after a downgrade from R390 to R384.

When nvidia-smi is run without any arguments to verify the installation, the following error message is displayed:

            

            
NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.
        

Workaround

Before uninstalling the R390 VIB, clear all parameters of the nvidia module to remove any registry keys that are available only for the R390 Virtual GPU Manager.

            

            
# esxcli system module parameters set -p "" -m nvidia
        

Status

Not an NVIDIA bug

Ref. #

200366884

5.59. Tesla P40 cannot be used in pass-through mode

Description

Pass-through mode on Tesla P40 GPUs and other GPUs based on the Pascal architecture does not work as expected. In some situations, after the VM is powered on, the guest OS crashes or fails to boot.

Workaround

Ensure that your GPUs are configured as described in Requirements for Using GPUs Requiring Large MMIO Space in Pass-Through Mode.

Status

Not a bug

Ref. #

1944539

5.60. On Linux, 3D applications run slowly when windows are dragged

Description

When windows for 3D applications on Linux are dragged, the frame rate drops substantially and the application runs slowly.

This issue does not affect 2D applications.

Status

Open

Ref. #

1949482

5.61. A segmentation fault in DBus code causes nvidia-gridd to exit on Red Hat Enterprise Linux and CentOS

Description

On Red Hat Enterprise Linux 6.8 and 6.9, and CentOS 6.8 and 6.9, a segmentation fault in DBus code causes the nvidia-gridd service to exit.

The nvidia-gridd service uses DBus for communication with NVIDIA X Server Settings to display licensing information through the Manage License page. Disabling the GUI for licensing resolves this issue.

To prevent this issue, the GUI for licensing is disabled by default. You might encounter this issue if you have enabled the GUI for licensing and are using Red Hat Enterprise Linux 6.8 or 6.9, or CentOS 6.8 and 6.9.

Version

Red Hat Enterprise Linux 6.8 and 6.9

CentOS 6.8 and 6.9

Status

Open

Ref. #

• 200358191
• 200319854
• 1895945

5.62. No Manage License option available in NVIDIA X Server Settings by default

Description

By default, the Manage License option is not available in NVIDIA X Server Settings. This option is missing because the GUI for licensing on Linux is disabled by default to work around the issue that is described in A segmentation fault in DBus code causes nvidia-gridd to exit on Red Hat Enterprise Linux and CentOS.

Workaround

This workaround requires sudo privileges.

If you are licensing a physical GPU for vCS, you must use the configuration file /etc/nvidia/gridd.conf.

1. If NVIDIA X Server Settings is running, shut it down.

2. If the /etc/nvidia/gridd.conf file does not already exist, create it by copying the supplied template file /etc/nvidia/gridd.conf.template.

3. As root, edit the /etc/nvidia/gridd.conf file to set the EnableUI option to TRUE.

4. Start the nvidia-gridd service.

   Copy

   Copied!

               
   
               
   # sudo service nvidia-gridd start
           

When NVIDIA X Server Settings is restarted, the Manage License option is now available.

Status

Open

5.63. Licenses remain checked out when VMs are forcibly powered off

Description

NVIDIA vGPU software licenses remain checked out on the license server when non-persistent VMs are forcibly powered off.

The NVIDIA service running in a VM returns checked out licenses when the VM is shut down. In environments where non-persistent licensed VMs are not cleanly shut down, licenses on the license server can become exhausted. For example, this issue can occur in automated test environments where VMs are frequently changing and are not guaranteed to be cleanly shut down. The licenses from such VMs remain checked out against their MAC address for seven days before they time out and become available to other VMs.

Resolution

If VMs are routinely being powered off without clean shutdown in your environment, you can avoid this issue by shortening the license borrow period. To shorten the license borrow period, set the LicenseInterval configuration setting in your VM image. For details, refer to Virtual GPU Client Licensing User Guide.

Status

Closed

Ref. #

1694975

5.64. Memory exhaustion can occur with vGPU profiles that have 512 Mbytes or less of frame buffer

Description

Memory exhaustion can occur with vGPU profiles that have 512 Mbytes or less of frame buffer.

This issue typically occurs in the following situations:

• Full screen 1080p video content is playing in a browser. In this situation, the session hangs and session reconnection fails.
• Multiple display heads are used with Citrix Virtual Apps and Desktops or VMware Horizon on a Windows 10 guest VM.
• Higher resolution monitors are used.
• Applications that are frame-buffer intensive are used.
• NVENC is in use.

To reduce the possibility of memory exhaustion, NVENC is disabled on profiles that have 512 Mbytes or less of frame buffer.

When memory exhaustion occurs, the NVIDIA host driver reports Xid error 31 and Xid error 43 in the VMware vSphere log file vmware.log in the guest VM’s storage directory.

The following vGPU profiles have 512 Mbytes or less of frame buffer:

• Tesla M6-0B, M6-0Q
• Tesla M10-0B, M10-0Q
• Tesla M60-0B, M60-0Q

The root cause is a known issue associated with changes to the way that recent Microsoft operating systems handle and allow access to overprovisioning messages and errors. If your systems are provisioned with enough frame buffer to support your use cases, you should not encounter these issues.

Workaround

• Use an appropriately sized vGPU to ensure that the frame buffer supplied to a VM through the vGPU is adequate for your workloads.
• Monitor your frame buffer usage.
• If you are using Windows 10, consider these workarounds and solutions:

  • Use a profile that has 1 Gbyte of frame buffer.

  • Optimize your Windows 10 resource usage.

    To obtain information about best practices for improved user experience using Windows 10 in virtual environments, complete the NVIDIA GRID vGPU Profile Sizing Guide for Windows 10 download request form.

    Additionally, you can use the VMware OS Optimization Tool to make and apply optimization recommendations for Windows 10 and other operating systems.

Status

Open

Ref. #

• 200130864
• 1803861

5.65. vGPU VM fails to boot in ESXi 6.5 if the graphics type is Shared

Description

On VMware vSphere Hypervisor (ESXi) 6.5, after vGPU is configured, VMs to which a vGPU is assigned may fail to start and the following error message may be displayed:

            

            
The amount of graphics resource available in the parent resource pool is insufficient for the operation.
        

The vGPU Manager VIB provides vSGA and vGPU functionality in a single VIB. After this VIB is installed, the default graphics type is Shared, which provides vSGA functionality. To enable vGPU support for VMs in VMware vSphere 6.5, you must change the default graphics type to Shared Direct. If you do not change the default graphics type you will encounter this issue.

Version

VMware vSphere Hypervisor (ESXi) 6.5

Workaround

Change the default graphics type to Shared Direct as explained in Virtual GPU Software User Guide.

Status

Open

Ref. #

200256224

5.66. ESXi 6.5 web client shows high memory usage even when VMs are idle

Description

On VMware vSphere Hypervisor (ESXi) 6.5, the web client shows a memory usage alarm with critical severity for VMs to which a vGPU is attached even when the VMs are idle. When memory usage is monitored from inside the VM, no memory usage alarm is shown. The web client does not show a memory usage alarm for the same VMs without an attached vGPU.

Version

VMware vSphere Hypervisor (ESXi) 6.5

Workaround

Avoid using the VMware vSphere Hypervisor (ESXi) 6.5 web client to monitor memory usage for VMs to which a vGPU is attached.

Status

Not an NVIDIA bug

Ref. #

200191065

5.67. NVIDIA driver installation may fail for VMs on a host in a VMware DRS cluster

Description

For VMware vSphere releases before 6.7 Update 1, the ESXi host on which VMs configured with NVIDIA vGPU reside must not be a member of an automated VMware Distributed Resource Scheduler (DRS) cluster. The installer for the NVIDIA driver for NVIDIA vGPU software cannot locate the NVIDIA vGPU software GPU card on a host in an automated VMware DRS Cluster. Any attempt to install the driver on a VM on a host in an automated DRS cluster fails with the following error:

            

            
NVIDIA Installer cannot continue
This graphics driver could not find compatible graphics hardware.
        

Version

VMware vSphere Hypervisor (ESXi) releases before 6.7 Update 1.

Workaround

Ensure that the automation level of the DRS cluster is set to Manual.

For more information about this setting, see Edit Cluster Settings in the VMware documentation.

Status

Open

Ref. #

1933449

5.68. GNOME Display Manager (GDM) fails to start on Red Hat Enterprise Linux 7.2 and CentOS 7.0

Description

GDM fails to start on Red Hat Enterprise Linux 7.2 and CentOS 7.0 with the following error:

            

            
Oh no! Something has gone wrong!
        

Workaround

Permanently enable permissive mode for Security Enhanced Linux (SELinux).

1. As root, edit the /etc/selinux/config file to set SELINUX to permissive.
   Copy

   Copied!

               
   
               
   SELINUX=permissive
           

2. Reboot the system.
   Copy

   Copied!

               
   
               
   ~]# reboot
           

For more information, see Permissive Mode in Red Hat Enterprise Linux 7 SELinux User's and Administrator's Guide.

Status

Not an NVIDIA bug

Ref. #

200167868

5.69. NVIDIA Control Panel fails to start and reports that “you are not currently using a display that is attached to an Nvidia GPU”

Description

When you launch NVIDIA Control Panel on a VM configured with vGPU, it fails to start and reports that you are not using a display attached to an NVIDIA GPU. This happens because Windows is using VMware’s SVGA device instead of NVIDIA vGPU.

Fix

Make NVIDIA vGPU the primary display adapter.

Use Windows screen resolution control panel to make the second display, identified as “2” and corresponding to NVIDIA vGPU, to be the active display and select the Show desktop only on 2 option. Click Apply to accept the configuration.

You may need to click on the Detect button for Windows to recognize the display connected to NVIDIA vGPU.

Status

Open

Ref. #

5.70. VM configured with more than one vGPU fails to initialize vGPU when booted

Description

Using the current VMware vCenter user interface, it is possible to configure a VM with more than one vGPU device. When booted, the VM boots in VMware SVGA mode and doesn’t load the NVIDIA driver. The additional vGPU devices are present in Windows Device Manager but display a warning sign, and the following device status:

            

            
Windows has stopped this device because it has reported problems. (Code 43)
        

Workaround

NVIDIA vGPU currently supports a single virtual GPU device per VM. Remove any additional vGPUs from the VM configuration before booting the VM.

Status

Open

Ref. #

5.71. A VM configured with both a vGPU and a passthrough GPU fails to start the passthrough GPU

Description

Using the current VMware vCenter user interface, it is possible to configure a VM with a vGPU device and a passthrough (direct path) GPU device. This is not a currently supported configuration for vGPU. The passthrough GPU appears in Windows Device Manager with a warning sign, and the following device status:

            

            
Windows has stopped this device because it has reported problems. (Code 43)
        

Workaround

Do not assign vGPU and passthrough GPUs to a VM simultaneously.

Status

Open

Ref. #

1735002

5.72. vGPU allocation policy fails when multiple VMs are started simultaneously

Description

If multiple VMs are started simultaneously, vSphere may not adhere to the placement policy currently in effect. For example, if the default placement policy (breadth-first) is in effect, and 4 physical GPUs are available with no resident vGPUs, then starting 4 VMs simultaneously should result in one vGPU on each GPU. In practice, more than one vGPU may end up resident on a GPU.

Workaround

Start VMs individually.

Status

Not an NVIDIA bug

Ref. #

200042690

5.73. Before Horizon agent is installed inside a VM, the Start menu’s sleep option is available

Description

When a VM is configured with a vGPU, the Sleep option remains available in the Windows Start menu. Sleep is not supported on vGPU and attempts to use it will lead to undefined behavior.

Workaround

Do not use Sleep with vGPU.

Installing the VMware Horizon agent will disable the Sleep option.

Status

Closed

Ref. #

200043405

5.74. vGPU-enabled VMs fail to start, nvidia-smi fails when VMs are configured with too high a proportion of the server’s memory.

Description

If vGPU-enabled VMs are assigned too high a proportion of the server’s total memory, the following errors occur:

• One or more of the VMs may fail to start with the following error:
  Copy

  Copied!

              
  
              
  The available Memory resources in the parent resource pool are insufficient for the operation
          

• When run in the host shell, the nvidia-smi utility returns this error:
  Copy

  Copied!

              
  
              
  -sh: can't fork
          

For example, on a server configured with 256G of memory, these errors may occur if vGPU-enabled VMs are assigned more than 243G of memory.

Workaround

Reduce the total amount of system memory assigned to the VMs.

Status

Closed

Ref. #

200060499

5.75. On reset or restart VMs fail to start with the error VMIOP: no graphics device is available for vGPU…

Description

On a system running a maximal configuration, that is, with the maximum number of vGPU VMs the server can support, some VMs might fail to start post a reset or restart operation.

Fix

Upgrade to ESXi 6.0 Update 1.

Status

Closed

Ref. #

200097546

5.76. nvidia-smi shows high GPU utilization for vGPU VMs with active Horizon sessions

Description

vGPU VMs with an active Horizon connection utilize a high percentage of the GPU on the ESXi host. The GPU utilization remains high for the duration of the Horizon session even if there are no active applications running on the VM.

Workaround

None

Status

Open

Partially resolved for Horizon 7.0.1:

• For Blast connections, GPU utilization is no longer high.
• For PCoIP connections, utilization remains high.

Ref. #

1735009

Notice

This document is provided for information purposes only and shall not be regarded as a warranty of a certain functionality, condition, or quality of a product. NVIDIA Corporation (“NVIDIA”) makes no representations or warranties, expressed or implied, as to the accuracy or completeness of the information contained in this document and assumes no responsibility for any errors contained herein. NVIDIA shall have no liability for the consequences or use of such information or for any infringement of patents or other rights of third parties that may result from its use. This document is not a commitment to develop, release, or deliver any Material (defined below), code, or functionality.

NVIDIA reserves the right to make corrections, modifications, enhancements, improvements, and any other changes to this document, at any time without notice.

Customer should obtain the latest relevant information before placing orders and should verify that such information is current and complete.

NVIDIA products are sold subject to the NVIDIA standard terms and conditions of sale supplied at the time of order acknowledgement, unless otherwise agreed in an individual sales agreement signed by authorized representatives of NVIDIA and customer (“Terms of Sale”). NVIDIA hereby expressly objects to applying any customer general terms and conditions with regards to the purchase of the NVIDIA product referenced in this document. No contractual obligations are formed either directly or indirectly by this document.

NVIDIA products are not designed, authorized, or warranted to be suitable for use in medical, military, aircraft, space, or life support equipment, nor in applications where failure or malfunction of the NVIDIA product can reasonably be expected to result in personal injury, death, or property or environmental damage. NVIDIA accepts no liability for inclusion and/or use of NVIDIA products in such equipment or applications and therefore such inclusion and/or use is at customer’s own risk.

NVIDIA makes no representation or warranty that products based on this document will be suitable for any specified use. Testing of all parameters of each product is not necessarily performed by NVIDIA. It is customer’s sole responsibility to evaluate and determine the applicability of any information contained in this document, ensure the product is suitable and fit for the application planned by customer, and perform the necessary testing for the application in order to avoid a default of the application or the product. Weaknesses in customer’s product designs may affect the quality and reliability of the NVIDIA product and may result in additional or different conditions and/or requirements beyond those contained in this document. NVIDIA accepts no liability related to any default, damage, costs, or problem which may be based on or attributable to: (i) the use of the NVIDIA product in any manner that is contrary to this document or (ii) customer product designs.

No license, either expressed or implied, is granted under any NVIDIA patent right, copyright, or other NVIDIA intellectual property right under this document. Information published by NVIDIA regarding third-party products or services does not constitute a license from NVIDIA to use such products or services or a warranty or endorsement thereof. Use of such information may require a license from a third party under the patents or other intellectual property rights of the third party, or a license from NVIDIA under the patents or other intellectual property rights of NVIDIA.

Reproduction of information in this document is permissible only if approved in advance by NVIDIA in writing, reproduced without alteration and in full compliance with all applicable export laws and regulations, and accompanied by all associated conditions, limitations, and notices.

THIS DOCUMENT AND ALL NVIDIA DESIGN SPECIFICATIONS, REFERENCE BOARDS, FILES, DRAWINGS, DIAGNOSTICS, LISTS, AND OTHER DOCUMENTS (TOGETHER AND SEPARATELY, “MATERIALS”) ARE BEING PROVIDED “AS IS.” NVIDIA MAKES NO WARRANTIES, EXPRESSED, IMPLIED, STATUTORY, OR OTHERWISE WITH RESPECT TO THE MATERIALS, AND EXPRESSLY DISCLAIMS ALL IMPLIED WARRANTIES OF NONINFRINGEMENT, MERCHANTABILITY, AND FITNESS FOR A PARTICULAR PURPOSE. TO THE EXTENT NOT PROHIBITED BY LAW, IN NO EVENT WILL NVIDIA BE LIABLE FOR ANY DAMAGES, INCLUDING WITHOUT LIMITATION ANY DIRECT, INDIRECT, SPECIAL, INCIDENTAL, PUNITIVE, OR CONSEQUENTIAL DAMAGES, HOWEVER CAUSED AND REGARDLESS OF THE THEORY OF LIABILITY, ARISING OUT OF ANY USE OF THIS DOCUMENT, EVEN IF NVIDIA HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. Notwithstanding any damages that customer might incur for any reason whatsoever, NVIDIA’s aggregate and cumulative liability towards customer for the products described herein shall be limited in accordance with the Terms of Sale for the product.

VESA DisplayPort

DisplayPort and DisplayPort Compliance Logo, DisplayPort Compliance Logo for Dual-mode Sources, and DisplayPort Compliance Logo for Active Cables are trademarks owned by the Video Electronics Standards Association in the United States and other countries.

HDMI

HDMI, the HDMI logo, and High-Definition Multimedia Interface are trademarks or registered trademarks of HDMI Licensing LLC.

OpenCL

OpenCL is a trademark of Apple Inc. used under license to the Khronos Group Inc.

Trademarks

NVIDIA, the NVIDIA logo, NVIDIA GRID, NVIDIA GRID vGPU, NVIDIA Maxwell, NVIDIA Pascal, NVIDIA Turing, NVIDIA Volta, GPUDirect, Quadro, and Tesla are trademarks or registered trademarks of NVIDIA Corporation in the U.S. and other countries. Other company and product names may be trademarks of the respective companies with which they are associated.