The documentation set for this product strives to use bias-free language. For the purposes of this documentation set, bias-free is defined as language that does not imply discrimination based on age, disability, gender, racial identity, ethnic identity, sexual orientation, socioeconomic status, and intersectionality. Exceptions may be present in the documentation due to language that is hardcoded in the user interfaces of the product software, language used based on RFP documentation, or language that is used by a referenced third-party product. Learn more about how Cisco is using Inclusive Language.
This appendix contains the following topics:
Table D-1 lists the minimum server firmware versions for the supported GPU cards.
|
|
---|---|
|
|
|
---|---|---|
PCIE 71 |
1.If you have a GPU card in slot 7, you cannot also have a Qlogic QLE2562 HBA card in slot 9. |
All supported GPU cards require enablement of the BIOS setting that allows greater than 4 GB of memory-mapped I/O (MMIO).
If the server is used in standalone mode, this BIOS setting is enabled by default:
Advanced > PCI Configuration > Memory Mapped I/O Above 4 GB [Enabled]
If you need to change this from a different setting, enter the BIOS Setup Utility by pressing F2 when prompted during bootup.
If the server is integrated with Cisco UCS Manager and controlled by a service profile, this setting is enabled by default in the service profile when a GPU is present.
To change this setting manually, use the following procedure.
Step 1 Refer to the Cisco UCS Manager configuration guide (GUI or CLI) for your release for instructions on configuring service profiles:
Cisco UCS Manager Configuration Guides
Step 2 Refer to the chapter on Configuring Server-Related Policies > Configuring BIOS Settings.
Step 3 In the section of your profile for PCI Configuration BIOS Settings, set Memory Mapped IO Above 4GB Config
to one of the following:
Note Cisco UCS Manager pushes BIOS configuration changes through a BIOS policy or default BIOS settings to the Cisco Integrated Management Controller (CIMC) buffer. These changes remain in the buffer and do not take effect until the server is rebooted.
In the table below, the cable that is used with the GPU is listed. It is also indicated whether the cable is included in the GPU BOM or must be ordered separately.
Separate = Cable must be ordered separately when the ordering tool prompts you.
Included = Cable is included with the GPU; no additional action is needed
|
|
|
|
---|---|---|---|
Step 1 Remove a GPU card (or a blank filler panel) from the PCIe riser assembly:
a. Shut down and power off the server as described in Shutting Down and Powering Off the Server.
b. Slide the server out the front of the rack far enough so that you can remove the top cover.
c. Remove the top cover as described in Removing or Replacing the Server Top Cover.
d. Lift the blue plastic retaining latch at the top of the PCIe riser until the latch is vertical (see Figure 3-20). The lever action disengages the riser’s connector from the motherboard socket.
e. Lift straight up on both ends of the PCIe riser and remove it from the server. Set the riser on an antistatic surface.
f. Open the hinged blue plastic card cover (see Figure 3-22).
g. Open the hinged card-tab retainer (see Figure 3-22). Pinch the two blue finger-grips toward the center and swing open the retainer.
h. Pull evenly on both corners of the GPU card to remove it from the socket on the PCIe riser.
i. Disconnect the GPU card’s power cable from the PCIe riser power connector.
a. Connect the white connector of the included power cable to the white power-connector on the riser. The power connector is adjacent to PCIE 2 on riser 1 or PCIE 6 on riser 2 (see Figure 3-22).
For a list of cables that go with supported double-wide GPUs, see Double-Wide GPU Power Cables.
Note Your GPU card might be shipped with two power cables: a straight cable and a Y-cable. The straight cable is used for connecting power to the GPU card in this server; do not use the Y-cable, which is used for connecting the GPU card in external devices only.
Note If your cable has a 10-pin connector, the white 10-pin connector on the cable connects to a white 8-pin socket on the riser. The cable connector is keyed so that when you insert it, only 8 pins are used. Two pins of the cable connector sit outside the riser socket (see Figure D-1.)
Figure D-1 High-Power GPU Power Cable (UCSC-300W-460M4) 10-Pin Connector
|
|
||
|
Two pins on cable connector that are outside the socket on riser (not used) |
|
b. Connect the black connector of the power cable to the black connector on the GPU card.
c. Align the new GPU card with the empty socket on the PCIe riser assembly (PCIE 2 on riser 1 or PCIE 7 on riser 2).
d. Push down evenly on both ends of the card until it is fully seated in the socket.
Ensure that the card’s rear-panel tab sits flat against the PCIe riser rear-panel opening.
e. Close the hinged card-tab retainer and press it down until it clicks and locks in place.
f. Close the hinged blue plastic card cover.
g. Align the riser so that its connector is over the motherboard socket.
h. With the blue plastic retaining latch fully open (vertical), lower the riser into the chassis alignment channels until its connector makes contact with the motherboard socket.
i. Close the retaining latch until it is flat to fully engage the riser with the motherboard socket.
k. Replace the server in the rack, replace cables, and then power on the server by pressing the Power button.
Step 3 If you installed a Tesla M60 or M10 GPU, continue with Using NVIDIA GRID License Server For M-Series GPUs.
This section applies to the following GPUs:
Use the topics in this appendix in the following order:
2. Register your product activation keys with NVIDIA.
3. Download the GRID software suite.
4. Install the GRID License Server software to a host.
5. Generate licenses on the NVIDIA Licensing Portal and download them.
7. Decide whether to use the GPU in compute mode or graphics mode.
The NVIDIA Tesla M-Series GPU combines Tesla and GRID functionality when the licensed GRID features such as GRID vGPU and GRID Virtual Workstation are enabled. These features are enabled during OS boot by borrowing a software license that is served over the network from the NVIDIA GRID License Server virtual appliance. The license is returned to the license server when the OS shuts down.
You obtain the licenses that are served by the GRID License Server from NVIDIA’s Licensing Portal as downloadable license files, which you install into the GRID License Server via its management interface (see Figure D-2).
Figure D-2 GRID Licensing Architecture
There are three editions of GRID licenses, which enable three different classes of GRID features. The GRID software automatically selects the license edition based on the features that you are using (see Table D-3 ).
|
|
---|---|
After your order is processed, NVIDIA sends you a Welcome email that contains your product activation keys (PAKs) and a list of the types and quantities of licenses that you purchased.
Step 1 Select the Log In link, or the Register link if you do not already have an account.
The NVIDIA Software Licensing Center > License Key Registration dialog opens.
Step 2 Complete the License Key Registration form and then click Submit My Registration Information.
The NVIDIA Software Licensing Center > Product Information Software dialog opens.
Step 3 If you have additional PAKs, click Register Additional Keys. For each additional key, complete the form on the License Key Registration dialog and then click Submit My Registration Information.
Step 4 Agree to the terms and conditions and set a password when prompted.
Step 1 Return to the NVIDIA Software Licensing Center > Product Information Software dialog.
Step 2 Click the Current Releases tab.
Step 3 Click the NVIDIA GRID link to access the Product Download dialog. This dialog includes download links for:
– NVIDIA License Manager software
Step 4 Use the links to download the software.
For full installation instructions and troubleshooting, refer to the NVIDIA GRID License Server User Guide. Also refer to the NVIDIA GRID License Server Release Notes for the latest information about your release.
The License Server requires a Java runtime environment and an Apache Tomcat installation. Apache Tomcat is installed when you use the NVIDIA installation wizard for Windows.
Step 1 Download and install the latest Java 32-bit runtime environment from https://www.oracle.com/downloads/index.html.
Note Install the 32-bit Java Runtime Environment, regardless of whether your platform is Windows 32-bit or 64-bit.
Step 2 Create a server interface:
a. On the NVIDIA Software Licensing Center dialog, click Grid Licensing > Create License Server.
b. On the Create Server dialog, fill in your desired server details.
c. Save the.bin file that is generated onto your license server for installation.
Step 3 Unzip the NVIDIA License Server installer Zip file that you downloaded previously and run setup.exe
.
Step 4 Accept the EULA for the NVIDIA License Server software and the Apache Tomcat software. Tomcat is installed automatically during the License Server installation.
Step 5 Use the installer wizard to step through the installation.
Note On the Choose Firewall Options dialog, select the ports to be opened in the firewall. NVIDIA recommends that you use the default setting, which opens port 7070 but leaves port 8080 closed.
Step 6 Verify the installation. Open a web browser on the License Server host and connect to the URL http://localhost:8080/licserver. If the installation was successful, you see the NVIDIA License Client Manager interface.
The License Server requires a Java runtime environment and an Apache Tomcat installation. You must install both separately before installing the License Server on Linux.
Step 1 Verify that Java was installed with your Linux installation. Use the following command:
If no Java version is displayed, use your Linux package manager to install with the following command:
Step 2 Use your Linux package manager to install the tomcat and tomcat-webapps packages.
a. Use the following command to install Tomcat:
b. Enable the Tomcat service for automatic startup on boot:
d. Verify that the Tomcat service is operational. Open a web browser on the License Server host and connect to the URL http://localhost:8080. If the installation was successful, you see the Tomcat webapp.
Step 3 Install the License Server:
a. Unpack the License Server tar file using the following command:
b. Run the unpacked setup binary as root:
c. Accept the EULA and then continue with the installation wizard to finish the installation.
Note On the Choose Firewall Options dialog, select the ports to be opened in the firewall. NVIDIA recommends that you use the default setting, which opens port 7070 but leaves port 8080 closed.
Step 4 Verify the installation. Open a web browser on the License Server host and connect to the URL http://localhost:8080/licserver. If the installation was successful, you see the NVIDIA License Client Manager interface.
Open a web browser on the License Server host and access the URL http://localhost:8080/licserver
If you configured the License Server host’s firewall to permit remote access to the License Server, the management interface is accessible from remote machines at the URL http://hostname:8080/licserver
Your License Server’s Ethernet MAC address is used as an identifier when registering the License Server with NVIDIA’s Licensing Portal.
Step 1 Access the GRID License Server Management Interface in a browser. See Accessing the GRID License Server Management Interface.
Step 2 In the left-side License Server panel, select Configuration. The License Server Configuration panel opens. Next to Server host ID, a pull-down menu lists the possible Ethernet MAC addresses.
Step 3 Select your License Server’s MAC address from the Server host ID pull-down.
Note It is important to use the same Ethernet ID consistently to identify the server when generating licenses on NVIDIA’s Licensing Portal. NVIDIA recommends that you select one entry for a primary, non-removable Ethernet interface on the platform.
Step 1 Access the GRID License Server Management Interface in a browser. See Accessing the GRID License Server Management Interface.
Step 2 In the left-side License Server panel, select Configuration. The License Server Configuration panel opens.
Step 3 Use the License Server Configuration menu to install the.bin file that you generated earlier.
b. Browse to the license.bin file that you want to install and click Open.
The license file is installed on your License Server. When installation is complete, you see the confirmation message, “Successfully applied license file to license server.”
Use the following procedure to view which licenses are installed and available, along with their properties.
Step 1 Access the GRID License Server Management Interface in a browser. See Accessing the GRID License Server Management Interface.
Step 2 In the left-side License Server panel, select Licensed Feature Usage.
Step 3 Click on a feature in the Feature column to see detailed information about the current usage of that feature.
Use the following procedure to view information about which licenses are currently in-use and borrowed from the server.
Step 1 Access the GRID License Server Management Interface in a browser. See Accessing the GRID License Server Management Interface.
Step 2 In the left-side License Server panel, select Licensed Clients.
Step 3 To view detailed information about a single licensed client, click on its Client ID in the list.
Features that require GRID licensing run at reduced capability until a GRID license is acquired.
To acquire a GRID license on Windows, use the following procedure.
Step 1 Open the NVIDIA Control Panel using one of the following methods:
Step 2 In the NVIDIA Control Panel left-pane under Licensing, select Manage License.
The Manage License task pane opens and shows the current license edition being used. The GRID software automatically selects the license edition based on the features that you are using. The default is Tesla (unlicensed).
Step 3 If you want to acquire a license for GRID Virtual Workstation, under License Edition, select GRID Virtual Workstation.
Step 4 In the License Server field, enter the address of your local GRID License Server.
The address can be a domain name or an IP address.
Step 5 In the Port Number field, enter your port number of leave it set to the default used by the server, which is 7070.
The system requests the appropriate license edition from your configured License Server. After a license is successfully acquired, the features of that license edition are enabled.
Note After you configure licensing settings in the NVIDIA Control Panel, the settings persist across reboots.
To acquire a GRID license on Linux, use the following procedure.
Step 1 Edit the configuration file /etc/nvidia/gridd.conf:
Step 2 Edit the ServerUrl
line with the address of your local GRID License Server.
The address can be a domain name or an IP address. See the example file below.
Step 3 Append the port number (default 7070) to the end of the address with a colon. See the example file below.
Step 4 Edit the FeatureType
line with the integer for the license type. See the example file below.
Step 5 Restart the nvidia-gridd service.
The service automatically acquires the license edition that you specified in the FeatureType line. You can confirm this in /var/log/messages
.
Note After you configure licensing settings in gridd.conf, the settings persist across reboots.
The NVIDIA Tesla M60 GPU is shipped in compute mode, which is optimized for high-performance compute (HPC) applications. However, while compute mode is best for HPC usage, it can cause compatibility issues with OS and hypervisors if you use the GPU primarily as a graphics device.
The mode is determined at power-on, from settings stored in the GPU’s non-volatile memory. You can use the command-line tool gpumodeswitch
to toggle the GPU between compute mode and graphics mode. Table D-4 and Table D-5 compare the compute mode and graphic mode default settings.
The command line utility gpumodeswitch
can be run in the following environments:
Note Consult NVIDIA product release notes for the latest information on compatibility with compute and graphic modes.
The gpumodeswitch utility supports the following commands:
This command writes information to a log file named listgpumodes.txt
in the current working directory.
Switches to graphics mode. Switches mode of all supported GPUs in the server unless you specify otherwise when prompted.
Switches to compute mode. Switches mode of all supported GPUs in the server unless you specify otherwise when prompted.
Note After you switch GPU mode, reboot the server to ensure that the modified resources of the GPU are correctly accounted for by any OS or hypervisor running on the server.
After you install the hardware, you must update to the correct level of server BIOS and then install GPU drivers and other software in this order:
Install the latest server BIOS by using the Host Upgrade Utility for the Cisco UCS C460 M4 server.
Note You must do this procedure before you update the GPU drivers.
Step 1 Navigate to the following URL: http://www.cisco.com/cisco/software/navigator.html.
Step 2 Click Servers–Unified Computing in the middle column.
Step 3 Click Cisco UCS C-Series Rack-Mount Standalone Server Software in the right-hand column.
Step 4 Click the name of your model of server in the right-hand column.
Step 5 Click Unified Computing System (UCS) Server Firmware.
Step 6 Click the release number.
Step 7 Click Download Now to download the ucs-server platform-huu-version_number.iso file.
Step 8 Verify the information on the next page, and then click Proceed With Download.
Step 9 Continue through the subsequent screens to accept the license agreement and browse to a location where you want to save the file.
Step 10 Use the Host Upgrade Utility to update the server BIOS.
The user guides for the Host Upgrade Utility are at Utility User Guides.
After you update the server BIOS, you can install GPU drivers to your hypervisor virtual machine.
Step 1 Install your hypervisor software on a computer. Refer to your hypervisor documentation for the installation instructions.
Step 2 Create a virtual machine in your hypervisor. Refer to your hypervisor documentation for instructions.
Step 3 Install the GPU drivers to the virtual machine. Download the drivers from either:
Step 5 Check that the virtual machine is able to recognize the GPU card. In Windows, use the Device Manager and look under Display Adapters.