Nvidia Grid Vgpu Profile Sizing For Windows

Nvidia grid vgpu profile sizing for windows

File Name: vgpu-profile-sizing-guidance-for-windows-10.pdf

File Size: 513.21 KB

File Type: Application/pdf

Last Modified: 1 year

Status: Available

Last checked: 2 days ago!

This Document Has Been Certified by a Professional

100% customizable

Language: English

We recommend downloading this file onto your computer


January 2017
NVIDIA Performance Engineering
Best Practices Guide
To provide the right level of user experience for your users, it’s important to provide
them with the right amount of resources based on their workloads. There are many
factors to consider when sizing a virtual desktop environment including CPU, RAM,
storage, network, and more. The purpose of this guide is to provide guidance on how to
assign the appropriate amount of frame buffer for Windows 10 knowledge worker use
cases in NVIDIA GRID™ environments

The latest operating system from Microsoft, Windows 10, was designed to deliver
improved user experience on both PCs and mobile devices. While Windows 10 has a
great look and feel, it also introduces more frequent OS updates. This higher frequency
of updates is driving many organizations to rethink how they manage their PC
refreshes. Desktop virtualization allows IT departments to more easily manage and
deploy these new upgrades, but new considerations must be made to accommodate the
user experience requirements. Graphics will become even more important when
Windows 10 is deployed in virtual environments to ensure that users get full
functionality and application compatibility

Windows 10 is proving to be the most graphics-intensive operating system that
Microsoft has ever released. As you move from a Windows 7 or 8 environment it’s
important to reevaluate the requirements of your virtual desktops as you transition

Frame buffer, or graphics memory, is a dedicated resource in NVIDIA GRID
deployments and can help you determine the overall density of your environment

NVIDIA GRID vGPU Profile Sizing for Windows 10 | 1
How to CHoose the Correct NVIDIA GRID vGPU Profile for Windows 10
Knowledge worker workloads will vary per user depending on many factors, including:
 Number of applications
 The types of applications
 File sizes
 Number of Monitors and their resolution
To size your vGPU profile correctly, all of these factors must be considered

To understand the impact of some of these factors on frame buffer usage, the NVIDIA
GRID Performance Engineering team conducted a number of tests using Windows 10
with the NVIDIA Tesla™ M10 GPU. Based on these tests, NVIDIA recommends that
users that have any of the following characteristics should be assigned a vPC 1 GB
profile to deliver optimal experience:
 Heavy application use; including WebGL, video streaming, and flash applications
 2560×1600 resolution or higher
 2 or more monitors
As with any workload, your results will vary. Therefore, NVIDIA recommends that you
conduct real user testing to get specific scalability numbers for your environment

To deliver optimal performance to your users, it’s important to look at several factors to
determine the correct vGPU profile. The guidance in the following sections is based on
testing by the NVIDIA GRID Performance Engineering team of different factors and
their effect on frame buffer usage. The suggested steps for testing within your own
environment are provided to help you get the most accurate results

Frame Buffer Usage per VM
Testing was conducted on two different VMs, both of which were configured with 2
vCPUs, 4 GB of vRAM, and NVIDIA GRID 4.1 software. The first VM had a Virtual PC
(vPC) license and 512 MB of frame buffer on a Tesla M10 (M10-0B profile).The second
VM also had a vPC license but the frame buffer was increased to 1 GB on the Tesla M10
(M10-1B profile)

NVIDIA GRID vGPU Profile Sizing for Windows 10 | 2
How to CHoose the Correct NVIDIA GRID vGPU Profile for Windows 10
Workload Resolution Monitors Frame buffer usage
M10-0B Profile
Windows 7 idle 1920x1080 (full HD) 1 168 MB
Windows 10 idle 1920x1080 (full HD) 1 229 MB
Windows 10 idle 2560x1600 1 287 MB
LoginVSI Win10 Knowledge Worker 1920x1080 (full HD) 1 411 MB
M10-1B Profile
Windows 10 idle 1920x1080 (full HD) 2 283 MB
LoginVSI Win10 Knowledge Worker 1920x1080 (full HD) 2 664 MB
Application Workload
A user’s application workload will always be the most important factor in determining
which profile will deliver optimal performance. Office workers today use many
applications simultaneously, often with multiple applications interacting with one
another. Based on a study with Lakeside Software, the number of applications that use
OpenGL or DirectX, and, therefore leverage graphics acceleration, has doubled in the
last 5 years. Today, almost 60% of users have at least one application that is graphics

To simulate application workloads in the testing, the LoginVSI Windows 10 Knowledge
Worker (medium) workload was used. Like any benchmark, this synthetic workload can
be used to simulate real user behavior but should not take the place of real user testing

To understand if users can use the Virtual PC edition with 512 MB of frame buffer (M10-
0B profile), it’s important to monitor frame buffer usage with a monitoring tool that
exposes GPU resources on their physical PCs or on their VMs. If usage exceeds 512 MB
of frame buffer, you should move to 1 GB of frame buffer (M10-1B profile)

A number of existing monitoring vendors have GPU metrics already built into their
platforms. If you do not currently use a monitoring solution, you can use free solutions
such as Microsoft Performance Monitor (PerfMon), NVIDIA System Management
Interface (nvidia-smi), GPUProfiler, and others

Monitor Resolution
As monitor resolutions continue to increase, more pixels are being delivered to the
screen. As a result, the frame buffer usage in a virtual environment increases. Today,
most users have full HD (1920×1080) resolution or above, which uses a minimum of 229
MB of frame buffer when Windows 10 is idle

NVIDIA GRID vGPU Profile Sizing for Windows 10 | 3
How to CHoose the Correct NVIDIA GRID vGPU Profile for Windows 10
While full HD is currently the most common resolution, an increasing number of
devices are being released with higher resolution screens. On a 2560×1600 monitor,
Windows 10 uses 287 MB of frame buffer when idle, about 60 MB more than on a
1920×1080 monitor

To provide enough frame buffer to support the application workload, NVIDIA
recommends using Virtual PC with 1 GB of frame buffer (M10-1B profile) when using
2560×1600 resolution for office workloads. To deliver higher resolutions, up to 4K
(4096×2160), you should use Virtual Workstation (vWS) with a minimum of 1 GB of
frame buffer (M10-1Q profile)

Number of Monitors
The number of monitors to which the VM is being delivered also affects frame buffer
usage. Today it is very common to see office workers with at least 2 monitors, which will
increase the frame buffer requirements

Testing showed that 2 full HD monitors running an idle Windows 10 VM used 283 MB
of frame buffer. When a workload was added on the two full HD monitors, frame buffer
usage reached 611 MB, which exceeds the 512 MB of the M10-0B profile

When you are supporting a knowledge worker environment that has two or more
monitors, NVIDIA requires that you use Virtual PC with 1 GB of frame buffer (M10-1B
profile). Depending on your overall workload, 3-4 monitors may require using Virtual
Workstation with 2 GB of frame buffer (M10-2Q) to deliver optimal performance

As with all scalability testing, user workloads and environment must be taken into
account for scalability analysis. In order to test NVIDIA GRID in your environment you
can choose to get started with a certified NVIDIA partner or start a proof of concept
(POC) with a certified server and the 90-day NVIDIA GRID evaluation license

Important things to remember during your POC
1. Define “acceptable” user experience

Defining user experience (UX) requires careful examination of user and application

A definition of acceptable user experience can be obvious, such as the rendering time
for an image to appear, or the ability to smoothly pan across that image. It can also
NVIDIA GRID vGPU Profile Sizing for Windows 10 | 4
How to CHoose the Correct NVIDIA GRID vGPU Profile for Windows 10
be less obvious, such as the ability to smoothly scroll down a page or the “snappy”
reaction for a menu to appear after a right click

To avoid generic feedback, ask users to report metrics, and to judge specific activities
or functions using finite scales (for example, 1-5, 5 being best)

2. Compare real-world workloads

In virtual environments, time-slicing of resources allows users to get the same level
of performance even when sharing resources

Time-slicing results from users’ thinking time, which includes any pause in their
interaction with the application, any period when they are not using the application,
or even sitting at their desks

By adding up all the time away from the application (for example, meetings, lunch,
and periods out of office) you could expect to get even more benefits from shared
resources. These benefits equate to more resources for each user’s session and
typically a more responsive application

A realistic estimate of user interaction with an application results in a better-
perceived experience by the end user than benchmarks based on peak workloads
with inhuman work that is unrepresentative of real users’ interactions with an

3. Test with real users

It’s important to actually look at the application running to be sure that the
experience is enjoyable for users

While idle systems or benchmarks can be used as a starting point, real user
workflows may require more or less graphics resources. Because the number of
monitors and their resolution will impact graphics requirements, users bringing
their own devices or logging in from different locations may impact the overall user

When you also consider the effect of real-world workloads, you can see why real
users are the most accurate means of testing

NVIDIA GRID vGPU Profile Sizing for Windows 10 | 5
The information provided in this specification is believed to be accurate and reliable as of the date provided. However, NVIDIA Corporation
(“NVIDIA”) does not give any representations or warranties, expressed or implied, as to the accuracy or completeness of such information

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 publication supersedes and replaces all other specifications for the product that may have been
previously supplied

NVIDIA reserves the right to make corrections, modifications, enhancements, improvements, and other changes to this specification, at
any time and/or to discontinue any product or service without notice. Customer should obtain the latest relevant specification 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. NVIDIA hereby
expressly objects to applying any customer general terms and conditions with regard to the purchase of the NVIDIA product referenced in
this specification

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 these specifications will be suitable for any specified use without
further testing or modification. Testing of all parameters of each product is not necessarily performed by NVIDIA. It is customer’s sole
responsibility to ensure the product is suitable and fit for the application planned by customer and to do 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 specification. NVIDIA does not accept any 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 specification, 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 specification. 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 specification is permissible only if reproduction is approved by
NVIDIA in writing, is reproduced without alteration, and is accompanied by all associated conditions, limitations, and notices

MERCHANTABILITY, AND FITNESS FOR A PARTICULAR PURPOSE. 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 NVIDIA terms and conditions 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, the HDMI logo, and High-Definition Multimedia Interface are trademarks or registered trademarks of HDMI Licensing LLC

ROVI Compliance Statement
NVIDIA Products that support Rovi Corporation’s Revision 7.1.L1 Anti-Copy Process (ACP) encoding technology can only be sold or
distributed to buyers with a valid and existing authorization from ROVI to purchase and incorporate the device into buyer’s products

This device is protected by U.S. patent numbers 6,516,132; 5,583,936; 6,836,549; 7,050,698; and 7,492,896 and other intellectual property
rights. The use of ROVI Corporation's copy protection technology in the device must be authorized by ROVI Corporation and is intended
for home and other limited pay-per-view uses only, unless otherwise authorized in writing by ROVI Corporation. Reverse engineering or
disassembly is prohibited

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

NVIDIA, the NVIDIA logo, NVIDIA GRID, and NVIDIA GRID vGPU are trademarks and/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

© 2016 NVIDIA Corporation. All rights reserved


NVIDIA GRID vGPU Profile Sizing for Windows 10 3 Workload Resolution Monitors Frame buffer usage M10-0B Profile Windows 7 idle 1920x1080 (full HD) 1 168 MB Windows 10 idle …

Download Now

Documemt Updated


Popular Download


Frequently Asked Questions

What is nvidia grid vgpu?

The NVIDIA GRID vGPU software package for VMware vSphere, consisting of the GRID Virtual GPU Manager for ESXi, and NVIDIA GRID vGPU drivers for Windows, 32- and 64-bit. VMware vSphere 2015 or later, obtainable from VMware. An installed Windows VM to be enabled with vGPU.

What are virtual gpu profiles?

Virtual GPU profiles determine the amount of frame buffer that can be allocated to your virtual machine. The vGPU profiles that are supported on NVIDIA GPU’s with NVIDIA GRID software, are the 1B (with 1024 MB of frame buffer) and 2B (with 2048 MB of frame buffer).

Which vgpu types support a single display with a fixed resolution?

These vGPU types support a single display with a fixed maximum resolution. A.1.6. NVIDIA A30 Virtual GPU Types This GPU supports MIG-backed virtual GPUs and time-sliced virtual GPUs.

What is the maximum number of vgpus per physical gpu?

Each vGPU is given a fixed share of the physical GPU's processing cycles, the amount of which depends on the vGPU type, which in turn determines the maximum number of vGPUs per physical GPU. For example, the maximum number of T4-4C vGPUs per physical GPU is 4.