SherbyNet - /page/wikibook/entry.mu

Micron Document

<!DOCTYPE html>

<title>GPU virtualization</title>

</head>

GPU virtualization

</h1>

</header>

GPU virtualization refers to technologies that allow the use of a <a href="Graphics_processing_unit" title="Graphics processing unit">GPU</a> to <a href="Hardware_acceleration" title="Hardware acceleration">accelerate</a> <a href="Computer_graphics" title="Computer graphics">graphics</a> or <a href="GPGPU" class="mw-redirect" title="GPGPU">GPGPU</a> applications running on a <a href="Virtual_machine" title="Virtual machine">virtual machine</a>. GPU virtualization is used in various applications such as <a href="Desktop_virtualization" title="Desktop virtualization">desktop virtualization</a>,<a href="#cite_note-vmware-gpu-virtualization-1">[1]</a> <a href="Cloud_gaming" title="Cloud gaming">cloud gaming</a><a href="#cite_note-gpu-consolidation-cloud-gaming-2">[2]</a> and <a href="Computational_science" title="Computational science">computational science</a> (e.g. <a href="Fluid_dynamics" title="Fluid dynamics">hydrodynamics</a> simulations).<a href="#cite_note-passthrough-performance-3">[3]</a>

GPU virtualization implementations generally involve one or more of the following techniques: device emulation, API remoting, fixed pass-through and mediated pass-through. Each technique presents different trade-offs regarding virtual machine to GPU <a href="Consolidation_ratio" title="Consolidation ratio">consolidation ratio</a>, <a href="Computer_graphics" title="Computer graphics">graphics</a> <a href="Hardware_acceleration" title="Hardware acceleration">acceleration</a>, rendering <a href="Image_quality" title="Image quality">fidelity</a> and <a href="Software_feature" title="Software feature">feature</a> support, <a href="Software_portability" title="Software portability">portability</a> to different hardware, <a href="Process_isolation" title="Process isolation">isolation</a> between virtual machines, and support for suspending/resuming and <a href="Live_migration" title="Live migration">live migration</a>.<a href="#cite_note-vmware-gpu-virtualization-1">[1]</a><a href="#cite_note-gpu-full-virtualization-workshop-4">[4]</a><a href="#cite_note-gvirt-5">[5]</a><a href="#cite_note-logv-6">[6]</a>

<div class="mw-heading mw-heading2"><h2 id="API_remoting">API remoting</h2></div>

In <a href="API" title="API">API</a> remoting or API forwarding, calls to graphical APIs from guest applications are forwarded to the host by <a href="Remote_procedure_call" title="Remote procedure call">remote procedure call</a>, and the host then executes graphical commands from multiple guests using the host's GPU as a single user.<a href="#cite_note-vmware-gpu-virtualization-1">[1]</a> It may be considered a form of <a href="Paravirtualization" class="mw-redirect" title="Paravirtualization">paravirtualization</a> when combined with device emulation.<a href="#cite_note-gpuvm-7">[7]</a> This technique allows sharing GPU resources between multiple guests and the host when the GPU does not support hardware-assisted virtualization. It is conceptually simple to implement, but it has several disadvantages:<a href="#cite_note-vmware-gpu-virtualization-1">[1]</a>

<ul><li>In pure API remoting, there is little isolation between virtual machines when accessing graphical APIs; isolation can be improved using paravirtualization</li>

<li>Performance ranges from 86% to as low as 12% of native performance in applications that issue a large number of drawing calls per <a href="Film_frame" title="Film frame">frame</a></li>

<li>A large number of API <a href="Subroutine" class="mw-redirect" title="Subroutine">entry points</a> must be forwarded, and partial implementation of entry points may decrease fidelity</li>

<li>Applications on guest machines may be limited to few available APIs</li></ul>

Hypervisors usually use <a href="Shared_memory" title="Shared memory">shared memory</a> between guest and host to maximize performance and minimize latency. Using a <a href="Virtual_network_interface" title="Virtual network interface">network interface</a> instead (a common approach in <a href="Parallel_rendering" title="Parallel rendering">distributed rendering</a>), third-party software can add support for specific APIs (e.g. <a href="RCUDA" title="RCUDA">rCUDA</a><a href="#cite_note-rcuda-8">[8]</a> for <a href="CUDA" title="CUDA">CUDA</a>) or add support for typical APIs (e.g. VMGL<a href="#cite_note-vmgl-9">[9]</a> for <a href="OpenGL" title="OpenGL">OpenGL</a>) when it is not supported by the hypervisor's software package, although <a href="Network_delay" title="Network delay">network delay</a> and <a href="Serialization" title="Serialization">serialization</a> <a href="Overhead_(computing)" title="Overhead (computing)">overhead</a> may outweigh the benefits.

<caption>Application support from API remoting virtualization technologies

</caption>

<th>Technology

</th>

<th><a href="Direct3D" title="Direct3D">Direct3D</a>

</th>

<th><a href="OpenGL" title="OpenGL">OpenGL</a>

</th>

<th><a href="Vulkan_(API)" class="mw-redirect" title="Vulkan (API)">Vulkan</a>

</th>

<th><a href="OpenCL" title="OpenCL">OpenCL</a>

</th>

</th></tr>

<tr>

<td><a href="VMware" title="VMware">VMware</a> Virtual Shared Graphics Acceleration (vSGA)<a href="#cite_note-vmware-hardware-accelerated-graphics-10">[10]</a><a href="#cite_note-11">[11]</a>

</td>

<td style="background:#FFB; color:black;vertical-align:middle;text-align:center;" class="table-partial">11

</td>

</td>

<td style="background:#9EFF9E;color:black;vertical-align:middle;text-align:center;" class="table-yes">Yes

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td></tr>

<tr>

<td><a href="Parallels_Desktop_for_Mac" title="Parallels Desktop for Mac">Parallels Desktop for Mac</a> 3D acceleration<a href="#cite_note-parallels-desktop-3d-acceleration-13">[13]</a>

</td>

</td>

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td></tr>

<tr>

<td><a href="Hyper-V" title="Hyper-V">Hyper-V</a> <a href="RemoteFX" title="RemoteFX">RemoteFX</a> vGPU<a href="#cite_note-hyperv-remotefx-vgpu-17">[15]</a><a href="#cite_note-18">[16]</a>

</td>

<td style="background:#9EFF9E;color:black;vertical-align:middle;text-align:center;" class="table-yes">12

</td>

<td style="background:#9EFF9E;color:black;vertical-align:middle;text-align:center;" class="table-yes">4.4

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#FFB; color:black;vertical-align:middle;text-align:center;" class="table-partial">1.1

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td></tr>

<tr>

<td><a href="VirtualBox" title="VirtualBox">VirtualBox</a> Guest Additions 3D driver<a href="#cite_note-virtualbox-graphics-19">[17]</a><a href="#cite_note-virtualbox-guest-additions-3d-acceleration-20">[18]</a><a href="#cite_note-phoronix-virtualbox-opengl-21">[19]</a>

</td>

</td>

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td></tr>

<tr>

<td>Thincast Workstation - Virtual 3D<a href="#cite_note-25">[21]</a>

</td>

<td style="background:#9EFF9E;color:black;vertical-align:middle;text-align:center;" class="table-yes">12.1

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#9EFF9E;color:black;vertical-align:middle;text-align:center;" class="table-yes">Yes

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td></tr>

<tr>

<td><a href="QEMU" title="QEMU">QEMU</a>/<a href="Kernel-based_Virtual_Machine" title="Kernel-based Virtual Machine">KVM</a> with Virgil 3D<a href="#cite_note-virgil3d-26">[22]</a><a href="#cite_note-lwn-virgil3d-27">[23]</a><a href="#cite_note-28">[24]</a><a href="#cite_note-29">[25]</a>

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#9EFF9E;color:black;vertical-align:middle;text-align:center;" class="table-yes">4.3

</td>

<td style="background: #E7E7FF; color:black; vertical-align: middle; text-align: center;" class="table-planned">Planned

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td></tr></tbody></table>

/* start https://en.wikipedia.org/ */

.mw-parser-output .reflist{margin-bottom:0.5em;list-style-type:decimal}@media screen{.mw-parser-output .reflist{font-size:90%}}.mw-parser-output .reflist .references{font-size:100%;margin-bottom:0;list-style-type:inherit}.mw-parser-output .reflist-columns-2{column-width:30em}.mw-parser-output .reflist-columns-3{column-width:25em}.mw-parser-output .reflist-columns{margin-top:0.3em}.mw-parser-output .reflist-columns ol{margin-top:0}.mw-parser-output .reflist-columns li{page-break-inside:avoid;break-inside:avoid-column}.mw-parser-output .reflist-upper-alpha{list-style-type:upper-alpha}.mw-parser-output .reflist-upper-roman{list-style-type:upper-roman}.mw-parser-output .reflist-lower-alpha{list-style-type:lower-alpha}.mw-parser-output .reflist-lower-greek{list-style-type:lower-greek}.mw-parser-output .reflist-lower-roman{list-style-type:lower-roman}

/* end https://en.wikipedia.org/ */

</style><div class="reflist reflist-upper-alpha">

<li id="cite_note-parallels-wined3d-15"><a href="#cite_ref-parallels-wined3d_15-0">^</a> Wrapped to OpenGL using <a href="Wine_(software)#Direct3D" title="Wine (software)">WineD3D</a>.<a href="#cite_note-togl-14">[14]</a>

</li>

<li id="cite_note-compatprofile-16"><a href="#cite_ref-compatprofile_16-0">^</a> Compatibility profile.

</li>

<li id="cite_note-virtualbox-wined3d-23"><a href="#cite_ref-virtualbox-wined3d_23-0">^</a> Experimental. Wrapped to OpenGL using <a href="Wine_(software)#Direct3D" title="Wine (software)">WineD3D</a>.<a href="#cite_note-virtualbox-d3d-22">[20]</a>

</li>

<li id="cite_note-experimental-24"><a href="#cite_ref-experimental_24-0">^</a> Experimental.

</li>

</ol></div></div>

<div class="mw-heading mw-heading2"><h2 id="Fixed_pass-through">Fixed pass-through</h2></div>

In fixed pass-through or GPU pass-through (a special case of <a href="Peripheral_Component_Interconnect" title="Peripheral Component Interconnect">PCI</a> pass-through), a GPU is accessed directly by a single virtual machine exclusively and permanently. This technique achieves 96–100% of native performance<a href="#cite_note-passthrough-performance-3">[3]</a> and high fidelity,<a href="#cite_note-vmware-gpu-virtualization-1">[1]</a> but the acceleration provided by the GPU cannot be shared between multiple virtual machines. As such, it has the lowest <a href="Consolidation_ratio" title="Consolidation ratio">consolidation ratio</a> and the highest cost, as each graphics-accelerated virtual machine requires an additional physical GPU.<a href="#cite_note-vmware-gpu-virtualization-1">[1]</a>

The following software technologies implement fixed pass-through:

<ul><li><a href="VMware" title="VMware">VMware</a> Virtual Dedicated Graphics Acceleration (vDGA)<a href="#cite_note-vmware-hardware-accelerated-graphics-10">[10]</a><a href="#cite_note-vmware-workstation-vsga-only-30">[a]</a></li>

<li><a href="Parallels_Workstation_Extreme" title="Parallels Workstation Extreme">Parallels Workstation Extreme</a><a href="#cite_note-parallels-workstation-extreme-dedicated-gpu-31">[26]</a></li>

<li><a href="Hyper-V" title="Hyper-V">Hyper-V</a> Discrete Device Assignment (DDA)<a href="#cite_note-hyperv-dda-32">[27]</a></li>

<li><a href="Citrix_Systems" title="Citrix Systems">Citrix</a> XenServer GPU pass-through<a href="#cite_note-citrix-hdx3dpro-33">[28]</a><a href="#cite_note-citrix-hypervisor-graphics-34">[29]</a></li>

<li><a href="Xen" title="Xen">Xen</a><a href="#cite_note-intel-gvtd-kvm-xen-35">[30]</a> and <a href="QEMU" title="QEMU">QEMU</a>/<a href="Kernel-based_Virtual_Machine" title="Kernel-based Virtual Machine">KVM</a><a href="#cite_note-intel-gvtd-kvm-xen-35">[30]</a><a href="#cite_note-phoronix-gvt-36">[31]</a> with <a href="Graphics_Virtualization_Technology" class="mw-redirect" title="Graphics Virtualization Technology">Intel GVT-d</a><a href="#cite_note-intel-gvt-37">[32]</a><a href="#cite_note-intel-gvt-update-38">[33]</a><a href="#cite_note-phoronix-gvt-36">[31]</a></li></ul>

<a href="VirtualBox" title="VirtualBox">VirtualBox</a> removed support for PCI pass-through in version 6.1.0.<a href="#cite_note-virtualbox-changelog-39">[34]</a>

For certain GPU models, Nvidia and AMD video card drivers attempt to detect the GPU is being accessed by a virtual machine and disable some or all GPU features.<a href="#cite_note-passthrough-detection-40">[35]</a> NVIDIA has recently changed virtualization rules for consumer GPUs by disabling the check in GeForce Game Ready driver 465.xx and later.<a href="#cite_note-41">[36]</a>

For NVIDIA, various architectures of desktop and laptop consumer GPUs can be passed through in various ways. For desktop graphics cards, passthrough can be done via the KVM using either the legacy or UEFI BIOS configuration via SeaBIOS and OVMF, respectively.

<div class="mw-heading mw-heading3"><h3 id="NVIDIA">NVIDIA</h3></div>

<div class="mw-heading mw-heading4"><h4 id="Desktops">Desktops</h4></div>

For desktops, most graphics cards can be passed through, although for graphics cards with the Pascal architecture or older, the VBIOS of the graphics card must be passed through in the virtual machine if the GPU is used to boot the host.<a href="#cite_note-:0-42">[37]</a>

<div class="mw-heading mw-heading4"><h4 id="Laptops">Laptops</h4></div>

For laptops, the NVIDIA driver checks for the presence of a battery via ACPI, and without a battery, an error will be returned. To avoid this, an acpitable created from text converted into Base64 is required to spoof a battery and bypass the check.<a href="#cite_note-:0-42">[37]</a>

<div class="mw-heading mw-heading5"><h5 id="Pascal_and_earlier">Pascal and earlier</h5></div>

For the laptop graphics cards that are Pascal and older, passthrough varies widely on the configuration of the graphics card. For laptops that do not have NVIDIA Optimus, such as the MXM variants, passthrough can be achieved through traditional methods. For laptops that have NVIDIA Optimus on as well as rendering through the CPU's integrated graphics framebuffer as opposed to its own, the passthrough is more complicated, requiring a remote rendering display or service, the use of Intel GVT-g, as well as integrating the VBIOS into the boot configuration due to the VBIOS being present in the laptop's system BIOS as opposed to the GPU itself. For laptops that have a GPU with NVIDIA Optimus and have a dedicated framebuffer, the configurations may vary. If NVIDIA Optimus can be switched off, then passthrough is possible through traditional means. However, if Optimus is the only configuration, then it is most likely that the VBIOS is present in the laptop's system BIOS, requiring the same steps as the laptop rendering only on the integrated graphics framebuffer, but an external monitor is also possible.<a href="#cite_note-43">[38]</a>

<div class="mw-heading mw-heading2"><h2 id="Mediated_pass-through">Mediated pass-through</h2></div>

In mediated device pass-through or full GPU virtualization, the GPU hardware provides <a href="Context_(computing)" title="Context (computing)">contexts</a> with <a href="Virtual_address_space" title="Virtual address space">virtual memory ranges</a> for each guest through <a href="Input%E2%80%93output_memory_management_unit" title="Input–output memory management unit">IOMMU</a> and the hypervisor sends graphical commands from guests directly to the GPU. This technique is a form of <a href="Hardware-assisted_virtualization" class="mw-redirect" title="Hardware-assisted virtualization">hardware-assisted virtualization</a> and achieves near-native<a href="#cite_note-mediated-performance-47">[b]</a> performance and high fidelity. If the hardware exposes contexts as full logical devices, then guests can use any API. Otherwise, APIs and drivers must manage the additional complexity of GPU contexts. As a disadvantage, there may be little isolation between virtual machines when accessing GPU resources.<a href="#cite_note-vmware-gpu-virtualization-1">[1]</a>

The following software and hardware technologies implement mediated pass-through:

<ul><li><a href="VMware" title="VMware">VMware</a> Virtual Shared Pass-Through Graphics Acceleration<a href="#cite_note-vmware-workstation-vsga-only-30">[a]</a> with Nvidia vGPU<a href="#cite_note-nvidia-vgpu-48">[42]</a> or AMD MxGPU<a href="#cite_note-amd-mxgpu-49">[43]</a></li>

<li><a href="Citrix_Systems" title="Citrix Systems">Citrix</a> XenServer shared GPU with Nvidia vGPU, AMD MxGPU or Intel GVT-g<a href="#cite_note-citrix-hdx3dpro-33">[28]</a><a href="#cite_note-citrix-hypervisor-graphics-34">[29]</a></li>

<li><a href="Xen" title="Xen">Xen</a><a href="#cite_note-xengt-50">[44]</a><a href="#cite_note-intel-gvtg-kvm-xen-51">[45]</a><a href="#cite_note-phoronix-gvt-36">[31]</a> and <a href="Kernel-based_Virtual_Machine" title="Kernel-based Virtual Machine">KVM</a><a href="#cite_note-kvmgt-52">[46]</a><a href="#cite_note-intel-gvtg-kvm-xen-51">[45]</a> with <a href="Graphics_Virtualization_Technology" class="mw-redirect" title="Graphics Virtualization Technology">Intel GVT-g</a><a href="#cite_note-intel-gvt-37">[32]</a><a href="#cite_note-intel-gvt-update-38">[33]</a></li>

<li>Thincast Workstation - Virtual 3D feature (Direct X 12 & Vulkan 3D API)</li></ul>

While API remoting is generally available for current and older GPUs, mediated pass-through requires hardware support available only on specific devices.

<caption>Hardware support for mediated pass-through virtualization

</caption>

<th rowspan="2">Vendor

</th>

<th rowspan="2">Technology

</th>

<th colspan="3"><a href="Graphics_processing_unit#Dedicated_graphics_processing_unit" title="Graphics processing unit">Dedicated graphics card</a> families

</th>

<th rowspan="2"><a href="Graphics_processing_unit#Integrated_graphics" title="Graphics processing unit">Integrated GPU</a> families

</th></tr>

<tr>

<th>Server

</th>

<th>Professional

</th>

<th>Consumer

</th></tr>

<tr>

<td><a href="Nvidia" title="Nvidia">Nvidia</a>

</td>

</td>

<td><a href="Nvidia_GRID" title="Nvidia GRID">GRID</a>, <a href="Nvidia_Tesla" title="Nvidia Tesla">Tesla</a>

</td>

<td><a href="Nvidia_Quadro" class="mw-redirect" title="Nvidia Quadro">Quadro</a>

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td data-sort-value="" style="vertical-align:middle; text-align:center" class="table-na">—

</td></tr>

<tr>

</td>

<td>MxGPU<a href="#cite_note-amd-mxgpu-49">[43]</a><a href="#cite_note-amd-firepro-mxgpu-54">[48]</a>

</td>

<td><a href="AMD_FirePro#FirePro_Server_Series_(S000x/Sxx_000)" title="AMD FirePro">FirePro Server</a>, <a href="Radeon_Instinct" class="mw-redirect" title="Radeon Instinct">Radeon Instinct</a>

</td>

<td><a href="Radeon_Pro" title="Radeon Pro">Radeon Pro</a>

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td>

<td style="background:#FFC7C7;color:black;vertical-align:middle;text-align:center;" class="table-no">No

</td></tr>

<tr>

<td><a href="Intel" title="Intel">Intel</a>

</td>

</td>

<td data-sort-value="" style="vertical-align:middle; text-align:center" class="table-na">—

</td>

<td data-sort-value="" style="vertical-align:middle; text-align:center" class="table-na">—

</td>

<td data-sort-value="" style="vertical-align:middle; text-align:center" class="table-na">—

</td>

<td><a href="Broadwell_(microarchitecture)" title="Broadwell (microarchitecture)">Broadwell</a> and newer

</td></tr></tbody></table>

<div class="mw-heading mw-heading2"><h2 id="Device_emulation">Device emulation</h2></div>

GPU architectures are very complex and change quickly, and their internal details are often kept secret. It is generally not feasible to fully virtualize new generations of GPUs, only older and simpler generations. For example, <a href="PCem" class="mw-redirect" title="PCem">PCem</a>, a specialized emulator of the <a href="IBM_PC_compatible" title="IBM PC compatible">IBM PC</a> architecture, can emulate a <a href="S3_ViRGE" title="S3 ViRGE">S3 ViRGE</a>/DX graphics device, which supports <a href="Direct3D" title="Direct3D">Direct3D</a> 3, and a <a href="Voodoo2" title="Voodoo2">3dfx Voodoo2</a>, which supports <a href="Glide_(API)" title="Glide (API)">Glide</a>, among others.<a href="#cite_note-pcem-graphics-55">[49]</a>

When using a <a href="Video_Graphics_Array" title="Video Graphics Array">VGA</a> or an <a href="Super_VGA" title="Super VGA">SVGA</a> virtual display adapter,<a href="#cite_note-vmware-svga-device-56">[50]</a><a href="#cite_note-virtualbox-svga-device-57">[51]</a><a href="#cite_note-qemu-svga-device-58">[52]</a> the guest may not have 3D graphics acceleration, providing only minimal functionality to allow access to the machine via a graphics terminal. The emulated device may expose only basic 2D graphics modes to guests. The virtual machine manager may also provide common API implementations using <a href="Software_rendering" title="Software rendering">software rendering</a> to enable 3D graphics applications on the guest, albeit at speeds that may be low as 3% of hardware-accelerated native performance.<a href="#cite_note-vmware-gpu-virtualization-1">[1]</a> The following software technologies implement graphics APIs using software rendering:

<ul><li><a href="VMware" title="VMware">VMware</a> SVGA 3D software renderer<a href="#cite_note-vmware-svga3d-59">[53]</a></li>

<li><a href="VirtualBox" title="VirtualBox">VirtualBox</a> VMSVGA graphics controller<a href="#cite_note-virtualbox-svga-device-57">[51]</a></li>

<li><a href="Citrix_Systems" title="Citrix Systems">Citrix</a> XenServer OpenGL Software Accelerator<a href="#cite_note-citrix-software-opengl-60">[54]</a></li>

<li><a href="Windows_Advanced_Rasterization_Platform" title="Windows Advanced Rasterization Platform">Windows Advanced Rasterization Platform</a></li>

<li><a href="Core_OpenGL" title="Core OpenGL">Core OpenGL</a> software renderer</li>

<li><a href="Mesa_(computer_graphics)" title="Mesa (computer graphics)">Mesa</a> software renderer</li>

<li>Swift Shader (implements <a href="WebGPU" title="WebGPU">WebGPU</a>) software renderer</li></ul>

<ul><li><a href="Comparison_of_platform_virtualization_software" title="Comparison of platform virtualization software">Comparison of platform virtualization software</a></li></ul>

<div class="mw-heading mw-heading2"><h2 id="Notes">Notes</h2></div>

<li id="cite_note-vmware-workstation-vsga-only-30">^ <a href="#cite_ref-vmware-workstation-vsga-only_30-0">a</a> <a href="#cite_ref-vmware-workstation-vsga-only_30-1">b</a> Not available on <a href="VMware_Workstation" title="VMware Workstation">VMware Workstation</a>.

</li>

<li id="cite_note-mediated-performance-47"><a href="#cite_ref-mediated-performance_47-0">^</a> Intel GVT-g achieves 80–90% of native performance.<a href="#cite_note-intel-gvtg-idf15-44">[39]</a><a href="#cite_note-intel-gvtg-xdc2017-45">[40]</a> Nvidia vGPU achieves 88–96% of native performance considering the overhead on a VMware hypervisor.<a href="#cite_note-vmware-vgpu-performance-46">[41]</a>

</li>

</ol></div></div>

<div class="mw-heading mw-heading2"><h2 id="References">References</h2></div>

/* start https://en.wikipedia.org/ */

.mw-parser-output cite.citation{font-style:inherit;word-wrap:break-word}.mw-parser-output .citation q{quotes:"\"""\"""'""'"}.mw-parser-output .citation:target{background-color:rgba(0,127,255,0.133)}.mw-parser-output .id-lock-free.id-lock-free a{background:url("./mw/Lock-green.svg")right 0.1em center/9px no-repeat}.mw-parser-output .id-lock-limited.id-lock-limited a,.mw-parser-output .id-lock-registration.id-lock-registration a{background:url("./mw/Lock-gray-alt-2.svg")right 0.1em center/9px no-repeat}.mw-parser-output .id-lock-subscription.id-lock-subscription a{background:url("./mw/Lock-red-alt-2.svg")right 0.1em center/9px no-repeat}.mw-parser-output .cs1-ws-icon a{background:url("./mw/Wikisource-logo.svg")right 0.1em center/12px no-repeat}body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .id-lock-free a,body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .id-lock-limited a,body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .id-lock-registration a,body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .id-lock-subscription a,body:not(.skin-timeless):not(.skin-minerva) .mw-parser-output .cs1-ws-icon a{background-size:contain;padding:0 1em 0 0}.mw-parser-output .cs1-code{color:inherit;background:inherit;border:none;padding:inherit}.mw-parser-output .cs1-hidden-error{display:none;color:var(--color-error,#d33)}.mw-parser-output .cs1-visible-error{color:var(--color-error,#d33)}.mw-parser-output .cs1-maint{display:none;color:#085;margin-left:0.3em}.mw-parser-output .cs1-kern-left{padding-left:0.2em}.mw-parser-output .cs1-kern-right{padding-right:0.2em}.mw-parser-output .citation .mw-selflink{font-weight:inherit}@media screen{.mw-parser-output .cs1-format{font-size:95%}html.skin-theme-clientpref-night .mw-parser-output .cs1-maint{color:#18911f}}@media screen and (prefers-color-scheme:dark){html.skin-theme-clientpref-os .mw-parser-output .cs1-maint{color:#18911f}}

/* end https://en.wikipedia.org/ */

</style><cite id="CITEREFDowtySugerman2009" class="citation journal cs1 cs1-prop-location-test">Dowty, Micah; Sugerman, Jeremy (July 2009). Written at <a href="San_Diego" title="San Diego">San Diego</a>. <a rel="nofollow" class="external text" href="https://www.usenix.org/legacy/event/wiov08/tech/full_papers/dowty/dowty.pdf">"GPU Virtualization on VMware's Hosted I/O Architecture"</a> (PDF). ACM SIGOPS Operating Systems Review. 43 (3). <a href="New_York_City" title="New York City">New York City</a>: <a href="Association_for_Computing_Machinery" title="Association for Computing Machinery">Association for Computing Machinery</a>: 73–82. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1145%2F1618525.1618534">10.1145/1618525.1618534</a>. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/0163-5980">0163-5980</a>. <a href="S2CID_(identifier)" class="mw-redirect" title="S2CID (identifier)">S2CID</a> <a rel="nofollow" class="external text" href="https://api.semanticscholar.org/CorpusID:228328">228328</a>. Retrieved 10 September 2020.</cite>

</li>

<li id="cite_note-gpu-consolidation-cloud-gaming-2"><a href="#cite_ref-gpu-consolidation-cloud-gaming_2-0">^</a> <cite id="CITEREFHongFan-ChiangLeeChen2014" class="citation conference cs1">Hong, Hua-Jun; Fan-Chiang, Tao-Ya; Lee, Che-Rung; Chen, Kuan-Ta; Huang, Chun-Ying; Hsu, Cheng-Hsin (2014). <a rel="nofollow" class="external text" href="https://www.iis.sinica.edu.tw/~swc/pub/gpu_virtualization_for_cloud_games.html">GPU Consolidation for Cloud Games: Are We There Yet?</a>. 13th Annual Workshop on Network and Systems Support for Games. <a href="Nagoya" title="Nagoya">Nagoya</a>: <a href="Institute_of_Electrical_and_Electronics_Engineers" title="Institute of Electrical and Electronics Engineers">Institute of Electrical and Electronics Engineers</a>. pp. 1–6. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2FNetGames.2014.7008969">10.1109/NetGames.2014.7008969</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-4799-6882-4</bdi>. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/2156-8138">2156-8138</a>. <a href="S2CID_(identifier)" class="mw-redirect" title="S2CID (identifier)">S2CID</a> <a rel="nofollow" class="external text" href="https://api.semanticscholar.org/CorpusID:664129">664129</a>. Retrieved 14 September 2020.</cite>

</li>

<li id="cite_note-passthrough-performance-3">^ <a href="#cite_ref-passthrough-performance_3-0">a</a> <a href="#cite_ref-passthrough-performance_3-1">b</a> <cite id="CITEREFWaltersYoungeKangYao2014" class="citation conference cs1">Walters, John; Younge, Andrew; Kang, Dong-In; Yao, Ke-Thia; Kang, Mikyung; Crago, Stephen; Fox, Geoffrey (2014). <a rel="nofollow" class="external text" href="https://ieeexplore.ieee.org/document/6973796">"GPU Passthrough Performance: A Comparison of KVM, Xen, VMware ESXi, and LXC for CUDA and OpenCL Applications"</a>. IEEE 7th International Conference on Cloud Computing. <a href="Anchorage" class="mw-redirect" title="Anchorage">Anchorage</a>: <a href="IEEE_Computer_Society" title="IEEE Computer Society">IEEE Computer Society</a>. pp. 636–643. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2FCLOUD.2014.90">10.1109/CLOUD.2014.90</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-4799-5063-8</bdi>. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/2159-6190">2159-6190</a>. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-gpu-full-virtualization-workshop-4"><a href="#cite_ref-gpu-full-virtualization-workshop_4-0">^</a> <cite id="CITEREFYuRossbach2017" class="citation conference cs1">Yu, Hangchen; Rossbach, Christopher (25 June 2017). <a rel="nofollow" class="external text" href="https://www.cs.utexas.edu/~hyu/publication/wddd17-gpuvm.pdf">Full Virtualization for GPUs Reconsidered</a> (PDF). <a href="International_Symposium_on_Computer_Architecture" title="International Symposium on Computer Architecture">ISCA</a>-44 14th Annual Workshop on Duplicating, Deconstructing and Debunking. <a href="Toronto" title="Toronto">Toronto</a>. Retrieved 12 September 2020.</cite>

</li>

<li id="cite_note-gvirt-5"><a href="#cite_ref-gvirt_5-0">^</a> <cite id="CITEREFTianDongCowperthwaite2014" class="citation conference cs1">Tian, Kun; Dong, Yaozu; Cowperthwaite, David (June 2014). <a rel="nofollow" class="external text" href="https://www.usenix.org/system/files/conference/atc14/atc14-paper-tian.pdf">"A Full GPU Virtualization Solution with Mediated Pass-Through"</a> (PDF). Proceedings of the 2014 USENIX Conference on USENIX Annual Technical Conference (USENIX ATC'14). <a href="USENIX_Annual_Technical_Conference" title="USENIX Annual Technical Conference">USENIX Annual Technical Conference</a>. <a href="Philadelphia" title="Philadelphia">Philadelphia</a>: <a href="USENIX" title="USENIX">USENIX</a>. pp. 121–132. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-931971-10-2</bdi>.</cite>

</li>

<li id="cite_note-logv-6"><a href="#cite_ref-logv_6-0">^</a> <cite id="CITEREFGottschlagHillenbrandKehneStoess2013" class="citation conference cs1">Gottschlag, Mathias; Hillenbrand, Marius; Kehne, Jens; Stoess, Jan; Bellosa, Frank (November 2013). <a rel="nofollow" class="external text" href="http://os.itec.kit.edu/downloads/logv_low_overhead%20gpgpu_virtualization.pdf">LoGV: Low-Overhead GPGPU Virtualization</a> (PDF). 10th <a href="International_Conference_on_High_Performance_Computing" title="International Conference on High Performance Computing">International Conference on High Performance Computing</a>. <a href="Zhangjiajie" title="Zhangjiajie">Zhangjiajie</a>: <a href="IEEE_Computer_Society" title="IEEE Computer Society">IEEE Computer Society</a>. pp. 1721–1726. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2FHPCC.and.EUC.2013.245">10.1109/HPCC.and.EUC.2013.245</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-0-7695-5088-6</bdi>. Retrieved 16 September 2020.</cite>

</li>

<li id="cite_note-gpuvm-7"><a href="#cite_ref-gpuvm_7-0">^</a> <cite id="CITEREFSuzukiKatoYamadaKono2014" class="citation conference cs1">Suzuki, Yusuke; Kato, Shinpei; Yamada, Hiroshi; Kono, Kenji (June 2014). <a rel="nofollow" class="external text" href="https://www.usenix.org/system/files/conference/atc14/atc14-paper-suzuki.pdf">"GPUvm: Why Not Virtualizing GPUs at the Hypervisor?"</a> (PDF). Proceedings of the 2014 USENIX Conference on USENIX Annual Technical Conference (USENIX ATC'14). <a href="USENIX_Annual_Technical_Conference" title="USENIX Annual Technical Conference">USENIX Annual Technical Conference</a>. <a href="Philadelphia" title="Philadelphia">Philadelphia</a>: <a href="USENIX" title="USENIX">USENIX</a>. pp. 109–120. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-931971-10-2</bdi>. Retrieved 14 September 2020.</cite>

</li>

<li id="cite_note-rcuda-8"><a href="#cite_ref-rcuda_8-0">^</a> <cite id="CITEREFDuatoPeñaSillaFernández2011" class="citation conference cs1">Duato, José; Peña, Antonio; Silla, Federico; Fernández, Juan; Mayo, Rafael; Quintana-Ortí, Enrique (December 2011). <a rel="nofollow" class="external text" href="https://core.ac.uk/download/pdf/231705177.pdf">Enabling CUDA acceleration within virtual machines using rCUDA</a> (PDF). 18th <a href="International_Conference_on_High_Performance_Computing" title="International Conference on High Performance Computing">International Conference on High Performance Computing</a>. <a href="Bangalore" class="mw-redirect" title="Bangalore">Bangalore</a>: <a href="IEEE_Computer_Society" title="IEEE Computer Society">IEEE Computer Society</a>. pp. 1–10. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1109%2FHiPC.2011.6152718">10.1109/HiPC.2011.6152718</a>. <a href="Hdl_(identifier)" class="mw-redirect" title="Hdl (identifier)">hdl</a>:<a rel="nofollow" class="external text" href="https://hdl.handle.net/2117%2F168226">2117/168226</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-4577-1951-6</bdi>. <a href="ISSN_(identifier)" class="mw-redirect" title="ISSN (identifier)">ISSN</a> <a rel="nofollow" class="external text" href="https://search.worldcat.org/issn/1094-7256">1094-7256</a>. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-vmgl-9"><a href="#cite_ref-vmgl_9-0">^</a> <cite id="CITEREFLagar-CavillaToliaSatyanarayananLara2007" class="citation conference cs1 cs1-prop-location-test">Lagar-Cavilla, Horacio; Tolia, Niraj; Satyanarayanan, Mahadev; Lara, Eyal (June 2007). <a rel="nofollow" class="external text" href="http://www.cs.cmu.edu/~satya/docdir/lagar-cavilla-vee-vmgl-2007.pdf">"VMM-Independent Graphics Acceleration"</a> (PDF). Written at <a href="San_Antonio" title="San Antonio">San Antonio</a>. Proceedings of the 3rd International Conference on Virtual Execution Environments. <a href="ACM_SIGOPS" title="ACM SIGOPS">VEE</a> '07. <a href="New_York_City" title="New York City">New York City</a>: <a href="Association_for_Computing_Machinery" title="Association for Computing Machinery">Association for Computing Machinery</a>. pp. 33–43. <a href="Doi_(identifier)" class="mw-redirect" title="Doi (identifier)">doi</a>:<a rel="nofollow" class="external text" href="https://doi.org/10.1145%2F1254810.1254816">10.1145/1254810.1254816</a>. <a href="ISBN_(identifier)" class="mw-redirect" title="ISBN (identifier)">ISBN</a> <bdi>978-1-59593-630-1</bdi>. Retrieved 12 September 2020.</cite>

</li>

<li id="cite_note-vmware-hardware-accelerated-graphics-10">^ <a href="#cite_ref-vmware-hardware-accelerated-graphics_10-0">a</a> <a href="#cite_ref-vmware-hardware-accelerated-graphics_10-1">b</a> <cite id="CITEREFLantinga" class="citation techreport cs1">Lantinga, Hilko. <a rel="nofollow" class="external text" href="https://techzone.vmware.com/resource/deploying-hardware-accelerated-graphics-vmware-horizon-7">Deploying Hardware-Accelerated Graphics with VMware Horizon</a> (Guide). <a href="VMware" title="VMware">VMware</a>. Retrieved 12 September 2020.</cite>

</li>

<li id="cite_note-11"><a href="#cite_ref-11">^</a> <cite id="CITEREFvisaac" class="citation web cs1">visaac. <a rel="nofollow" class="external text" href="https://docs.vmware.com/en/VMware-Workstation-Pro/16/rn/VMware-Workstation-16-Pro-Release-Notes.html">"VMware Workstation 16 Pro Release Notes"</a>. docs.vmware.com. Retrieved 2021-03-24.</cite>

</li>

<li id="cite_note-12"><a href="#cite_ref-12">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://www.phoronix.com/news/SVGA-Mesa-22-OpenGL-4.3">"VMware's SVGA Gallium3D Driver Lands OpenGL 4.3 Support In Mesa 22.0"</a>.</cite>

</li>

<li id="cite_note-parallels-desktop-3d-acceleration-13"><a href="#cite_ref-parallels-desktop-3d-acceleration_13-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://download.parallels.com/desktop/v16/docs/en_US/Parallels%20Desktop%20User's%20Guide/43132.htm">"Graphics Settings"</a>. <a rel="nofollow" class="external text" href="https://download.parallels.com/desktop/v16/docs/en_US/Parallels%20Desktop%20User's%20Guide/">Parallels Desktop - User's Guide</a> (Guide). <a href="Parallels_(company)" title="Parallels (company)">Parallels</a>.</cite>

</li>

<li id="cite_note-togl-14"><a href="#cite_ref-togl_14-0">^</a> <cite id="CITEREFBright2014" class="citation news cs1">Bright, Peter (11 March 2014). <a rel="nofollow" class="external text" href="https://arstechnica.com/gaming/2014/03/valve-releases-open-source-direct3d-to-opengl-translator/">"Valve releases open source Direct3D to OpenGL translator"</a>. <a href="Ars_Technica" title="Ars Technica">Ars Technica</a>. Retrieved 15 September 2020.</cite>

</li>

<li id="cite_note-hyperv-remotefx-vgpu-17"><a href="#cite_ref-hyperv-remotefx-vgpu_17-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://docs.microsoft.com/en-us/windows-server/virtualization/hyper-v/deploy/deploy-graphics-devices-using-remotefx-vgpu">"Deploy graphics devices using RemoteFX vGPU"</a>. <a rel="nofollow" class="external text" href="https://docs.microsoft.com/en-us/windows-server/virtualization/hyper-v/hyper-v-on-windows-server">Hyper-V on Windows Server</a> (Manual). <a href="Microsoft" title="Microsoft">Microsoft</a>. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-18"><a href="#cite_ref-18">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://docs.microsoft.com/en-us/windows-server/virtualization/hyper-v/plan/plan-for-gpu-acceleration-in-windows-server">"Plan for GPU acceleration in Windows Server"</a>. <a rel="nofollow" class="external text" href="https://docs.microsoft.com/en-us/windows-server/virtualization/hyper-v/hyper-v-on-windows-server">Hyper-V on Windows Server</a> (Manual). <a href="Microsoft" title="Microsoft">Microsoft</a>. Retrieved 15 September 2020.</cite>

</li>

<li id="cite_note-virtualbox-graphics-19"><a href="#cite_ref-virtualbox-graphics_19-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://docs.oracle.com/en/virtualization/virtualbox/6.1/user/guestadd-video.html">"Hardware-Accelerated Graphics"</a>. <a rel="nofollow" class="external text" href="https://docs.oracle.com/en/virtualization/virtualbox/6.1/user/index.html">Oracle VM VirtualBox User Manual</a> (Manual). <a href="Oracle_Corporation" title="Oracle Corporation">Oracle Corporation</a>. Retrieved 12 September 2012.</cite>

</li>

<li id="cite_note-virtualbox-guest-additions-3d-acceleration-20"><a href="#cite_ref-virtualbox-guest-additions-3d-acceleration_20-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://www.virtualbox.org/manual/ch04.html">"Guest Additions"</a>. <a rel="nofollow" class="external text" href="https://www.virtualbox.org/manual/">Oracle VM VirtualBox User Manual</a> (Manual). <a href="Oracle_Corporation" title="Oracle Corporation">Oracle Corporation</a>. Retrieved 12 September 2020.</cite>

</li>

<li id="cite_note-phoronix-virtualbox-opengl-21"><a href="#cite_ref-phoronix-virtualbox-opengl_21-0">^</a> <cite id="CITEREFLarabel2018" class="citation news cs1">Larabel, Michael (19 December 2018). <a rel="nofollow" class="external text" href="https://www.phoronix.com/scan.php?page=article&item=virtualbox-60-vmsvga&num=1">"VirtualBox 6.0 3D/OpenGL Performance With VMSVGA Adapter"</a>. <a href="Phoronix_Test_Suite" title="Phoronix Test Suite">Phoronix</a>. Retrieved 15 September 2020.</cite>

</li>

<li id="cite_note-virtualbox-d3d-22"><a href="#cite_ref-virtualbox-d3d_22-0">^</a> <cite id="CITEREFLarabel2009" class="citation news cs1">Larabel, Michael (29 January 2009). <a rel="nofollow" class="external text" href="https://www.phoronix.com/scan.php?page=news_item&px=NzAyNA">"VirtualBox Gets Accelerated Direct3D Support"</a>. <a href="Phoronix_Test_Suite" title="Phoronix Test Suite">Phoronix</a>. Retrieved 15 September 2020.</cite>

</li>

<li id="cite_note-25"><a href="#cite_ref-25">^</a> <a rel="nofollow" class="external text" href="https://www.freerdp.com/2019/02/20/hi-thincast-workstation">Hi! - The Thincast Workstation</a> FreeRDP Blog

</li>

<li id="cite_note-virgil3d-26"><a href="#cite_ref-virgil3d_26-0">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://virgil3d.github.io/">"Virgil 3D GPU project"</a>. <a href="GitHub" title="GitHub">GitHub</a> (Project). <a href="Freedesktop.org" title="Freedesktop.org">freedesktop.org</a>. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-lwn-virgil3d-27"><a href="#cite_ref-lwn-virgil3d_27-0">^</a> <cite id="CITEREFEdge2014" class="citation techreport cs1">Edge, Jake (10 September 2014). <a rel="nofollow" class="external text" href="https://lwn.net/Articles/611212/">Virgil 3D: A virtual GPU</a> (Article). <a href="LWN.net" title="LWN.net">LWN.net</a>. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-28"><a href="#cite_ref-28">^</a> <cite id="CITEREFWollny2019" class="citation news cs1">Wollny, Gert (28 August 2019). <a rel="nofollow" class="external text" href="https://www.collabora.com/news-and-blog/blog/2019/08/28/virglrenderer-state-of-virtualized-virtual-worlds/">"Virglrenderer and the state of virtualized virtual worlds"</a>. <a href="Collabora" title="Collabora">Collabora</a> News & Blog. Retrieved 15 September 2020.</cite>

</li>

<li id="cite_note-29"><a href="#cite_ref-29">^</a> <cite id="CITEREFHoffmann2019" class="citation news cs1">Hoffmann, Gerd (28 November 2019). <a rel="nofollow" class="external text" href="https://www.kraxel.org/blog/2019/11/virtio-gpu-status-and-plans/">"virtio gpu status and plans"</a>. Retrieved 15 September 2020.</cite>

</li>

<li id="cite_note-parallels-workstation-extreme-dedicated-gpu-31"><a href="#cite_ref-parallels-workstation-extreme-dedicated-gpu_31-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="http://download.parallels.com/doc/pwe/en/GPU_development_solution_brief.pdf">GPU Development with Parallels Workstation Extreme</a> (PDF) (White paper). <a href="Parallels_(company)" title="Parallels (company)">Parallels</a>. 2010. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-hyperv-dda-32"><a href="#cite_ref-hyperv-dda_32-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://docs.microsoft.com/en-us/windows-server/virtualization/hyper-v/deploy/deploying-graphics-devices-using-dda">"Deploy graphics devices using Discrete Device Assignment"</a>. <a rel="nofollow" class="external text" href="https://docs.microsoft.com/en-us/windows-server/virtualization/hyper-v/hyper-v-on-windows-server">Hyper-V on Windows Server</a> (Manual). <a href="Microsoft" title="Microsoft">Microsoft</a>. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-citrix-hdx3dpro-33">^ <a href="#cite_ref-citrix-hdx3dpro_33-0">a</a> <a href="#cite_ref-citrix-hdx3dpro_33-1">b</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://docs.citrix.com/en-us/xenapp-and-xendesktop/7-15-ltsr/graphics/hdx-3d-pro.html">"HDX 3D Pro"</a>. <a rel="nofollow" class="external text" href="https://docs.citrix.com/en-us/xenapp-and-xendesktop/7-15-ltsr">XenApp and XenDesktop 7.15 LTSR</a> (Manual). <a href="Citrix_Systems" title="Citrix Systems">Citrix Systems</a>. Retrieved 15 September 2020.</cite>

</li>

<li id="cite_note-citrix-hypervisor-graphics-34">^ <a href="#cite_ref-citrix-hypervisor-graphics_34-0">a</a> <a href="#cite_ref-citrix-hypervisor-graphics_34-1">b</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://docs.citrix.com/en-us/citrix-hypervisor/graphics.html">"Graphics overview"</a>. <a rel="nofollow" class="external text" href="https://docs.citrix.com/en-us/citrix-hypervisor">Citrix Hypervisor 8.2</a> (Manual). <a href="Citrix_Systems" title="Citrix Systems">Citrix Systems</a>. Retrieved 15 September 2020.</cite>

</li>

<li id="cite_note-intel-gvtd-kvm-xen-35">^ <a href="#cite_ref-intel-gvtd-kvm-xen_35-0">a</a> <a href="#cite_ref-intel-gvtd-kvm-xen_35-1">b</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://github.com/intel/gvt-linux/wiki/GVTd_Setup_Guide">GVT-d Setup Guide</a>. <a href="GitHub" title="GitHub">GitHub</a> (Guide). Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-phoronix-gvt-36">^ <a href="#cite_ref-phoronix-gvt_36-0">a</a> <a href="#cite_ref-phoronix-gvt_36-1">b</a> <a href="#cite_ref-phoronix-gvt_36-2">c</a> <cite id="CITEREFLarabel2014" class="citation news cs1">Larabel, Michael (4 May 2014). <a rel="nofollow" class="external text" href="https://www.phoronix.com/scan.php?page=news_item&px=MTY4MTc">"Intel Pushes Their Graphics Virtualization Capabilities"</a>. <a href="Phoronix_Test_Suite" title="Phoronix Test Suite">Phoronix</a>. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-intel-gvt-37">^ <a href="#cite_ref-intel-gvt_37-0">a</a> <a href="#cite_ref-intel-gvt_37-1">b</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://01.org/sites/default/files/documentation/gvt_flyer_final.pdf">"Bringing New Use Cases and Workloads to the Cloud with Intel Graphics Virtualization Technology (Intel GVT-g)"</a> (PDF). Intel Open Source Technology Center (Flyer). <a href="Intel" title="Intel">Intel</a>. 2016. Retrieved 14 August 2020.</cite>

</li>

<li id="cite_note-intel-gvt-update-38">^ <a href="#cite_ref-intel-gvt-update_38-0">a</a> <a href="#cite_ref-intel-gvt-update_38-1">b</a> <cite id="CITEREFJain2014" class="citation web cs1">Jain, Sunil (4 May 2014). <a rel="nofollow" class="external text" href="https://01.org/blogs/2014/intel%C2%AE-graphics-virtualization-update">"Intel Graphics Virtualization Update"</a> (Article). <a href="Intel" title="Intel">Intel</a>. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-virtualbox-changelog-39"><a href="#cite_ref-virtualbox-changelog_39-0">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://www.virtualbox.org/wiki/Changelog-6.1">"Changelog for VirtualBox 6.1"</a>. <a href="VirtualBox" title="VirtualBox">VirtualBox</a>. <a href="Oracle_Corporation" title="Oracle Corporation">Oracle Corporation</a>. 10 December 2019. Retrieved 12 September 2020.</cite>

</li>

<li id="cite_note-passthrough-detection-40"><a href="#cite_ref-passthrough-detection_40-0">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://wiki.archlinux.org/index.php/PCI_passthrough_via_OVMF#Video_card_driver_virtualisation_detection">"PCI passthrough via OVMF - Video card driver virtualization detection"</a>. <a href="Arch_Linux" title="Arch Linux">Arch Linux</a> Wiki (Wiki). Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-41"><a href="#cite_ref-41">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://nvidia.custhelp.com/app/answers/detail/a_id/5173/~/geforce-gpu-passthrough-for-windows-virtual-machine-%28beta%29">"GeForce GPU Passthrough for Windows Virtual Machine (Beta)"</a>. NVIDIA Support. 2021-03-30.</cite>

</li>

<li id="cite_note-:0-42">^ <a href="#cite_ref-:0_42-0">a</a> <a href="#cite_ref-:0_42-1">b</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://wiki.archlinux.org/title/PCI_passthrough_via_OVMF#Video_card_driver_virtualisation_detection">"PCI passthrough via OVMF - ArchWiki"</a>. wiki.archlinux.org. Retrieved 2021-05-20.</cite>

</li>

<li id="cite_note-43"><a href="#cite_ref-43">^</a> <cite id="CITEREFTian2020" class="citation web cs1">Tian, Lan (2020-06-25). <a rel="nofollow" class="external text" href="https://lantian.pub/en/article/modify-computer/laptop-intel-nvidia-optimus-passthrough.lantian/#Stop-Host-OS-from-Tampering-with-NVIDIA-GPU">"Intel and NVIDIA GPU Passthrough on a Optimus MUXless Laptop"</a>.</cite>

</li>

<li id="cite_note-intel-gvtg-idf15-44"><a href="#cite_ref-intel-gvtg-idf15_44-0">^</a> <cite id="CITEREFZheng2015" class="citation conference cs1">Zheng, Xiao (August 2015). <a rel="nofollow" class="external text" href="https://01.org/sites/default/files/documentation/sz15_sfts002_100_engf.pdf">Media Cloud Based on Intel Graphics Virtualization Technology (Intel GVT-g) and OpenStack</a> (PDF). <a href="Intel_Developer_Forum" title="Intel Developer Forum">Intel Developer Forum</a> (Presentation slide). <a href="San_Francisco" title="San Francisco">San Francisco</a>: <a href="Intel" title="Intel">Intel</a>. Retrieved 14 September 2020.</cite>

</li>

<li id="cite_note-intel-gvtg-xdc2017-45"><a href="#cite_ref-intel-gvtg-xdc2017_45-0">^</a> <cite id="CITEREFWang2017" class="citation conference cs1">Wang, Zhenyu (September 2017). <a rel="nofollow" class="external text" href="https://www.x.org/wiki/Events/XDC2017/wang_gvt.pdf">Full GPU virtualization in mediated pass-through way</a> (PDF). <a rel="nofollow" class="external text" href="https://www.x.org/wiki/Events/XDC2017/">XDC2017</a> (Presentation slide). <a href="Mountain_View%2C_California" title="Mountain View, California">Mountain View, California</a>: <a href="X.Org_Foundation" title="X.Org Foundation">X.Org Foundation</a>. Retrieved 14 September 2020.</cite>

</li>

<li id="cite_note-vmware-vgpu-performance-46"><a href="#cite_ref-vmware-vgpu-performance_46-0">^</a> <cite id="CITEREFKurkure2017" class="citation techreport cs1">Kurkure, Uday (12 October 2017). <a rel="nofollow" class="external text" href="https://blogs.vmware.com/performance/2017/10/episode-3-performance-comparison-native-gpu-virtualized-gpu-scalability-virtualized-gpus-machine-learning.html">Performance Comparison of Native GPU to Virtualized GPU and Scalability of Virtualized GPUs for Machine Learning</a>. VMware VROOM! Performance Blog (Article). <a href="VMware" title="VMware">VMware</a>. Episode 3. Retrieved 14 September 2020.</cite>

</li>

<li id="cite_note-nvidia-vgpu-48"><a href="#cite_ref-nvidia-vgpu_48-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://docs.nvidia.com/grid/latest/grid-vgpu-user-guide/">Virtual GPU Software User Guide</a> (Guide). <a href="Nvidia" title="Nvidia">Nvidia</a>. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-amd-mxgpu-49">^ <a href="#cite_ref-amd-mxgpu_49-0">a</a> <a href="#cite_ref-amd-mxgpu_49-1">b</a> <cite id="CITEREFWong2016" class="citation techreport cs1">Wong, Tonny (28 January 2016). <a rel="nofollow" class="external text" href="https://www.amd.com/system/files/documents/amd-mxgpu-white-paper.pdf">AMD multiuser GPU: hardware-enabled GPU virtualization for a true workstation experience</a> (PDF) (White paper). <a href="AMD" title="AMD">AMD</a>. Retrieved 12 September 2020.</cite>

</li>

<li id="cite_note-xengt-50"><a href="#cite_ref-xengt_50-0">^</a> <cite id="CITEREFWang2018" class="citation pressrelease cs1">Wang, Hongbo (18 October 2018). <a rel="nofollow" class="external text" href="https://01.org/igvt-g/blogs/wangbo85/2018/2018-q3-release-xengt-intel-gvt-g-xen">"2018-Q3 release of XenGT (Intel GVT-g for Xen)"</a> (Press release). <a href="Intel_Developer_Zone" title="Intel Developer Zone">Intel Open Source Technology Center</a>. Retrieved 14 August 2020.</cite>

</li>

<li id="cite_note-intel-gvtg-kvm-xen-51">^ <a href="#cite_ref-intel-gvtg-kvm-xen_51-0">a</a> <a href="#cite_ref-intel-gvtg-kvm-xen_51-1">b</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://github.com/intel/gvt-linux/wiki/GVTg_Setup_Guide">GVT-g Setup Guide</a>. <a href="GitHub" title="GitHub">GitHub</a> (Guide). Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-kvmgt-52"><a href="#cite_ref-kvmgt_52-0">^</a> <cite id="CITEREFWang2018" class="citation pressrelease cs1">Wang, Hongbo (18 October 2018). <a rel="nofollow" class="external text" href="https://01.org/igvt-g/blogs/wangbo85/2018/2018-q3-release-kvmgt-intel-gvt-g-kvm">"2018-Q3 release of KVMGT (Intel GVT-g for KVM)"</a> (Press release). <a href="Intel_Developer_Zone" title="Intel Developer Zone">Intel Open Source Technology Center</a>. Retrieved 14 August 2020.</cite>

</li>

<li id="cite_note-nvidia-virtualization-support-53"><a href="#cite_ref-nvidia-virtualization-support_53-0">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://docs.nvidia.com/grid/gpus-supported-by-vgpu.html">"NVIDIA Virtual GPU Software Supported GPUs"</a>. <a href="Nvidia" title="Nvidia">Nvidia</a>. Retrieved 9 September 2020.</cite>

</li>

<li id="cite_note-amd-firepro-mxgpu-54"><a href="#cite_ref-amd-firepro-mxgpu_54-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://www.amd.com/system/files/documents/firepro-s-series-datasheet.pdf">AMD FirePro S-Series for Virtualization</a> (PDF) (Datasheet). <a href="AMD" title="AMD">AMD</a>. 2016. Retrieved 13 September 2020.</cite>

</li>

<li id="cite_note-pcem-graphics-55"><a href="#cite_ref-pcem-graphics_55-0">^</a> <cite class="citation web cs1"><a rel="nofollow" class="external text" href="https://pcem-emulator.co.uk/status.html">"Systems/motherboards emulated"</a>. PCem (Project). Retrieved 26 October 2020.</cite>

</li>

<li id="cite_note-vmware-svga-device-56"><a href="#cite_ref-vmware-svga-device_56-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://docs.vmware.com/en/VMware-Tools/10.1.0/com.vmware.vsphere.vmwaretools.doc/GUID-6994A5F9-B62B-4BF1-99D8-E325874A4C7A.html?hWord=N4IghgNiBcIGoFkDuYBOBTABAFQPa4gGdMBlOAcQEEQBfIA">"VMware Tools Device Drivers"</a>. <a rel="nofollow" class="external text" href="https://docs.vmware.com/en/VMware-Tools/index.html">VMware Tools Documentation</a> (Manual). <a href="VMware" title="VMware">VMware</a>. Retrieved 12 September 2020.</cite>

</li>

<li id="cite_note-virtualbox-svga-device-57">^ <a href="#cite_ref-virtualbox-svga-device_57-0">a</a> <a href="#cite_ref-virtualbox-svga-device_57-1">b</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://www.virtualbox.org/manual/ch03.html">"Configuring Virtual Machines"</a>. <a rel="nofollow" class="external text" href="https://www.virtualbox.org/manual/">Oracle VM VirtualBox User Manual</a> (Manual). <a href="Oracle_Corporation" title="Oracle Corporation">Oracle Corporation</a>. Retrieved 12 September 2020.</cite>

</li>

<li id="cite_note-qemu-svga-device-58"><a href="#cite_ref-qemu-svga-device_58-0">^</a> <cite class="citation techreport cs1">"Display options". <a rel="nofollow" class="external text" href="https://www.qemu.org/docs/master/qemu-doc.html">QEMU User Documentation</a>. <a href="QEMU" title="QEMU">QEMU</a> (Manual). Retrieved 12 September 2020.</cite>

</li>

<li id="cite_note-vmware-svga3d-59"><a href="#cite_ref-vmware-svga3d_59-0">^</a> <cite id="CITEREFLong2013" class="citation techreport cs1">Long, Simon (2013). <a rel="nofollow" class="external text" href="https://www.vmware.com/content/dam/digitalmarketing/vmware/en/pdf/whitepaper/vmware-horizon-view-graphics-acceleration-deployment-white-paper.pdf">Virtual Machine Graphics Acceleration Deployment Guide</a> (PDF) (White paper). <a href="VMware" title="VMware">VMware</a>. Retrieved 14 September 2020.</cite>

</li>

<li id="cite_note-citrix-software-opengl-60"><a href="#cite_ref-citrix-software-opengl_60-0">^</a> <cite class="citation techreport cs1"><a rel="nofollow" class="external text" href="https://docs.citrix.com/en-us/xenapp-and-xendesktop/7-15-ltsr/graphics/opengl-software-accelerator.html">"OpenGL Software Accelerator"</a>. <a rel="nofollow" class="external text" href="https://docs.citrix.com/en-us/xenapp-and-xendesktop/7-15-ltsr">XenApp and XenDesktop 7.15 LTSR</a> (Manual). <a href="Citrix_Systems" title="Citrix Systems">Citrix Systems</a>. Retrieved 15 September 2020.</cite>

</li>

</ol></div></div></div><div><div class="zim-footer">

This article is issued from <a class="external text" title="Last edited on 2025-06-25" href="https://en.wikipedia.org/wiki/?title=GPU_virtualization&oldid=1297248669">Wikipedia</a>. The text is available under <a class="external text" href="https://creativecommons.org/licenses/by-sa/4.0/deed.en">Creative Commons Attribution-Share Alike 4.0</a> unless otherwise noted. Additional terms may apply for the media files.

</div>

</div></div>

</div>

</main>

</div>

</body></html>