For general computing and graphics acceleration scenarios, GPU-accelerated instances can provide better computing and graphics rendering capabilities only after drivers are installed on the instances. You can configure parameters to load an NVIDIA GRID driver when you create a GPU-accelerated instance. You can also install a GRID driver by using Cloud Assistant after you create a GPU-accelerated instance. This topic describes how to automatically load a GRID driver by using a community image pre-installed with the driver when you create a GPU-accelerated instance.
Procedure
Free community images pre-installed with GRID drivers are provided on the Community Images tab. You can search for the required community image by image ID and use the image to load the GRID driver.
Go to the instance buy page in the Elastic Compute Service (ECS) console.
Click the Custom Launch tab.
Configure settings for the instance based on your business requirements. The settings include the billing method, region, zone, instance type, and image.
This section describes how to configure the instance type and image. For more information about other settings, see Parameter settings. The following table lists the IDs of community images that are pre-installed with GRID drivers for GPU-accelerated instances in different regions.
NoteIf your instance resides in a region that is not included in the following table, submit a ticket to obtain the IDs of the community images that are pre-installed with GRID drivers in the region. The GRID driver version must match the instance type and region of your GPU-accelerated instance. Therefore, you must provide the information about your instance, such as the instance type, OS, and region when you submit a ticket.
vGPU-accelerated Linux instance
Region
Driver version
Community image name
Community image ID
Japan (Tokyo)
GRID 13.5
AliLinux2.1903_SWL
m-6we7hvjs7vemf2xvty3j
AliLinux3.2104_SWL
m-6weh121p8sreq859zax9
CentOS7.9_SWL
m-6we0wqwvjv70mkn8z8pm
Ubuntu_18.04_SWL
m-6we76maddq2qygmuv76s
Ubuntu_20.04_SWL
m-6webzle7xko46u27vpju
Ubuntu_22.04_SWL
m-6we7hvjs7vemf0yupw4e
Singapore
GRID 13.5
AliLinux2.1903_SWL
m-t4nd0i3ti9337eqgcfse
AliLinux3.2104_SWL
m-t4n7rgku461dtoaq2wtw
CentOS7.9_SWL
m-t4n6rm3s7h1ayx9cn00n
Ubuntu_18.04_SWL
m-t4napvpzvi1xulgz978b
Ubuntu_20.04_SWL
m-t4nf5uqofn0vvqjracjy
Ubuntu_22.04_SWL
m-t4n8lkv4jzoqb0me4y6m
Germany (Frankfurt)
GRID 13.5
AliLinux2.1903_SWL
m-gw83ebuzccacxrkigs5v
AliLinux3.2104_SWL
m-gw8gfwq0t7x9xo5obu93
CentOS7.9_SWL
m-gw88gpj79i07fy5u0jt9
Ubuntu_18.04_SWL
m-gw8atq53lgzc899upqo2
Ubuntu_20.04_SWL
m-gw81u32f6i1vdzu4jxs9
Ubuntu_22.04_SWL
m-gw8a476uh6dcrejffgsw
China (Hong Kong)
GRID 13.5
AliLinux2.1903_SWL
m-j6ccnjmjogi1lk8n4ua3
AliLinux3.2104_SWL
m-j6ciceb38trw6mwpsr26
CentOS7.9_SWL
m-j6cbt3t0005gthviwqee
Ubuntu_18.04_SWL
m-j6c8ij3p9mx1nbmntart
Ubuntu_20.04_SWL
m-j6c8pg9a7gvobq25aa3r
Ubuntu_22.04_SWL
m-j6c8ij3p9mx1nbmntars
Indonesia (Jakarta)
GRID 13.5
AliLinux2.1903_SWL
m-k1a3dydsfrzfiefruut5
AliLinux3.2104_SWL
m-k1a2sm5kvgj44fvduo7w
CentOS7.9_SWL
m-k1a19u2ycqayqxfza8cz
Ubuntu_18.04_SWL
m-k1a19u2ycqayqxfza8cy
Ubuntu_20.04_SWL
m-k1a40e8fvu72lh25ikji
Ubuntu_22.04_SWL
m-k1a28a2igiikv1597jhy
In the Image section, click the Community Image tab and enter the community image ID that you obtained.
The following figure shows an example on how to enter the ID of a community image. In this example, the community image of an
ecs.sgn7i-vws-m2s.xlargeinstance that runs Ubuntu in the Germany (Frankfurt) region is used.
GPU-accelerated compute-optimized or vGPU-accelerated Windows instance
Region
Driver version
Community image name
Community image ID
Japan (Tokyo)
GRID 13.5
windows_server_2019_SWL_13.5
m-6weazl3lpyj1ahn7gjge
windows_server_2022_SWL_13.5
m-6we3a1irnys18yuzq76j
GRID 15.2
windows_server_2019_SWL_15.2
m-6web9k0cruot3fspyh53
windows_server_2022_SWL_15.2
m-6wed2tia0dmbvm18rt3h
Singapore
GRID 13.5
windows_server_2019_SWL_13.5
m-t4n3w9p3ql1ywgabhz62
windows_server_2022_SWL_13.5
m-t4n6in4jq9an686rpgc8
GRID 15.2
windows_server_2019_SWL_15.2
m-t4ncmk1prn31xmvrqjbm
windows_server_2022_SWL_15.2
m-t4n57fn54e6zxnb7d5z5
Germany (Frankfurt)
GRID 13.5
windows_server_2019_SWL_13.5
m-gw8bh5qbxu0uw8yae51u
windows_server_2022_SWL_13.5
m-gw844y143c6an8h6hbsb
GRID 15.2
windows_server_2019_SWL_15.2
m-gw81tm68pk4bi98sjj5r
windows_server_2022_SWL_15.2
m-gw83ebuzccacxtjjku50
China (Hong Kong)
GRID 13.5
windows_server_2019_SWL_13.5
m-j6cd75kd3bf26gpplxy4
windows_server_2022_SWL_13.5
m-j6cgqrjxzn0tmckduj6o
GRID 15.2
windows_server_2019_SWL_15.2
m-j6ccwjdgry2qklgeulrp
windows_server_2022_SWL_15.2
m-j6cgjlxu873ht5tfqqu1
Indonesia (Jakarta)
GRID 13.5
windows_server_2019_SWL_13.5
m-k1a19u2ycqayqzf0eac4
windows_server_2022_SWL_13.5
m-k1afezw92fp5k869o06o
GRID 15.2
windows_server_2019_SWL_15.2
m-k1a43pskdpasb7kqljf0
windows_server_2022_SWL_15.2
m-k1a28a2igiikux76zfjp
In the Image section, click the Community Image tab and enter the community image ID that you obtained. The following figure shows an example on how to enter the ID of a community image. In this example, the community image of an
ecs.sgn7i-vws-m2s.xlargeinstance that runs Windows in the Germany (Frankfurt) region is used.
Follow the on-screen instructions to complete the payment.
References
If the system does not automatically load a GRID driver when you create a GPU-accelerated compute-optimized instance for general computing and graphics acceleration scenarios, you must separately install the driver after the instance is created. For more information, see the following topics: