immers.cloud is a cloud GPU platform that lets you host virtual servers equipped with powerful NVIDIA GPUs: ideal for everything from cloud gaming and 3D rendering to machine learning and running large language and generative AI models.
We offer pay-as-you-go-billing, meaning you pay only for the exact time your server is running.
The Sign-up process takes seconds: no long forms, no approvals, no waiting. Create your first GPU instance and start working within minutes of registration.
Our GPU servers are used for:
When you host a GPU server, you get full, dedicated access to the hardware. Everything is managed through an intuitive web dashboard or via OpenStack API.
Read moreThe immers.cloud platform enforces quotas and limits:
Your current quotas are displayed when you click on your username in the top menu.
Default quotas for all new users:
Current quotas can be increased upon request — simply contact technical support.
Platform limits on immers.cloud:
The difference lies in the payment method. Individuals pay for services using a personal bank card, while businesses can request invoices and make payments via bank transfer.
An individual user need to finish registration, confirm their email address, and top up their account balance.
A business organization representative needs to finish registration, confirm the corporate email address, add the organization and pay the first invoice.
You can change your password in your profile settings.
Before creating a Virtual Machine, top up your account balance to the required amount.
Next, create a Key Pair. Once the Key Pair is ready, you can immediately proceed to create a Virtual Machine by clicking the "Create Virtual Machine" button, or do so later in the Virtual Machines section.
Then:
The Virtual Machine creation process will begin. Wait until the status changes to Active.
Your Virtual Machine is now ready for use.
Important: Please note that the selected operating system will be installed automatically, but any additional software must be installed manually.
Volume-backed is a type of Virtual Machine where the boot disk data is stored on distributed network storage using HDD drives. Cinder volumes are used as boot disks.
Important: When creating a Volume-backed Virtual Machine, a volume is automatically created.
For Local instance type Virtual Machines, storage is located on local SSD drives.
Accordingly, the cost of disk storage for Volume-backed Virtual Machines is calculated according to HDD rates, while for Local instance type Virtual Machines, it is calculated according to SSD rates.
Note that Backup, Create Snapshot, Rescue, and Rebuild operations are available only for Virtual Machines of the Local type.
Comparison of performance speed between Volume-backed and Local instance types
By default, we provide connection via Remote Desktop Protocol (RDP). If you need to connect to the Virtual Machine using other methods, you must configure them manually.
For cloud images, standard usernames such as ubuntu, debian, fedora, opensuse, or cloud-user are typically used.
The SSH client is installed by default on Windows Server 2019, Windows 10 version 1809, and later builds.
ssh -i keyname-username.pem username@ip_hostFor example: ssh -i C:\Users\YourUserName\keyname-username.pem ubuntu@194.135.112.207.
keyname-username.pem.chmod 600 /full/path/to/file/keyname-username.pemssh -i /full/path/to/file/keyname-username.pem -p 22 username@111.11.111.111 Where:For cloud images, standard usernames such as ubuntu, debian, fedora, opensuse, or cloud-user are typically used.
chmod 600 keyname-username.pemThis ensures the private key file is readable and writable only by its owner.
Use the following command in the terminal to connect: ssh -i /path/to/key/file username@ip_host
To shut down a Virtual Machine, use the Stop command in the Actions dropdown menu.
Important: Billing for vCPU and RAM is stopped only when the Virtual Machine is shut down using this method.
Changing the configuration is available for Virtual Machines with the status Active or Shutoff.
In the Actions dropdown menu, select Resize. Choose a new configuration and click the Change Configuration button.
Important: For Volume-backed Virtual Machines, the disk size will not change when resizing the instance. To increase the disk space, you must separately expand the Volume.
Wait until the status changes to Verify Resize.
To confirm the changes, go to the Actions dropdown menu and select Confirm Resize. To cancel the changes, select Revert Resize.
The configuration change is complete. Wait for the Virtual Machine status to update.
Video tutorial
Yes, it is possible.
For Volume-backed Virtual Machines, you need to expand the volume:
For Local Virtual Machines, use the Resize option in the Actions dropdown menu and select a configuration with a larger disk size.
We continue billing for stopped Virtual Machines only for reserved resources: SSD and HDD disks (based on their size), external IP addresses, personal VLANs, and reserved PCI-Express devices (available only for GPU servers). vCPU and RAM are not billed when the machine is stopped.
To avoid charges for an inactive Virtual Machine, you must either delete it or archive it (Shelve).
If you have a Volume-backed Virtual Machine, after deletion the associated Volume will remain in the Volumes section. This Volume contains all your data and can be used later to recreate the exact same Virtual Machine. If you do not plan to reuse this Volume, you should delete it manually — otherwise, it will continue to incur storage charges.
Important: Before deleting a Volume-backed Virtual Machine, make sure to stop it using the Stop command in the Actions dropdown menu. Failure to do so may affect the integrity and usability of the Volume.
Note: You can deploy only one Virtual Machine from a single Volume. If you need to create multiple Virtual Machines from the same source, create a Volume Snapshot instead. From one Snapshot, you can launch any number of Virtual Machines.
If you have a Local Virtual Machine, create a Snapshot before deletion. When you need to use the Virtual Machineagain, simply redeploy it from that Snapshot.
Alternatively, you can use the Shelve option in the Actions menu. Shelve archives the server and releases compute resources, so you are no longer charged for CPU, RAM, or disk usage. You will only be billed for the storage of the server’s Snapshot (or Volume) and its IP address. A Snapshot is created automatically when you use Shelve.
Note: While the server is in the shelved_offloaded state, you cannot detach the Volume from it. To resume operation, select Unshelve from the Actions menu.
Costs for Snapshot and Volume storage are available on the Pricing page.
Download and install the FileZilla program.
Download the private key PPK file from the Key Pairs section. The file has the format keyname-username.pem.ppk.
Set up a server connection in FileZilla:
You can now download the required files.
We do not block any ports on Virtual Machines. All settings related to opening or closing ports are configured within the guest operating system. For Virtual Machines with an external IP address, all ports open in the operating system will be accessible from the outside.
You can configure a standard Virtual Machine once and then create any number of copies. The cloning process depends on the instance type.
This snapshot can now be selected when creating a new Virtual Machine. All Virtual Machines created from it will be exact copies of the original.
Get instant access to server creation: without submitting requests or waiting for approvals.. Your virtual machine is ready almost immediately after ordering. Resource allocation and OS installation happen automatically and usually take just a few minutes, so you can start working within minutes of registration.
Before deleting a virtual machine, you must shut it down by using the Stop command in the Actions dropdown menu. Once the virtual machine status changes to Shutoff, you can delete it or free up its resources (Shelve command).
To learn how to avoid being charged for a powered-off virtual machine, read here.
Important: Before deleting a Local-type virtual machine, create a snapshot if you plan to reuse its data later.
There is no minimum hosting period for servers on our platform.
We offer pay-as-you-go billing, so you only pay for the actual time your virtual machine is running. This gives you instant access to the latest GPU accelerators for any tasks, whether it takes minutes or months.
Use your virtual machine for as long as you need.
Currently, only one Availability Zone is available, located in the data center in Moscow, Russia.
ImmersView is a professional, low-latency screen streaming solution integrated directly with the immers.cloud platform — Russian alternative to Parsec and Sunshine.
Download and install the ImmersView Client on your local computer.
Launch the application and log in using your immers.cloud account credentials.
If you haven’t created a cloud server yet, you can do so either through the ImmersView client or via your immers.cloud dashboard.
Download and install the ImmersView Server component on your cloud instance.
Important: Reboot the server after installation to ensure proper operation.
In the ImmersView client, select the cloud server you want to connect to.
On the login screen, move your cursor to the top of the ImmersView window and click the Password button.
This will open a password decryption page on immers.cloud. Copy the displayed password and paste it into the login field.
ImmersView delivers interactive desktop streaming at up to 2K resolution and 120 Hz, with:
For more details, visit the official ImmersView project page.
After a configuration change is completed, Virtual Machines enter the VERIFY_RESIZE status. You must confirm that the resize was successful by selecting Confirm Resize in the Actions dropdown menu. If you wish to cancel the change, use the Revert Resize option instead.
Press Windows + R to open the Run dialog, type gpedit.msc, and press Enter.
Navigate to the following policy settings related to RDP:
Local Computer Policy\Computer Configuration\Administrative Templates\Windows Components\Remote Desktop Services\Remote Desktop Session Host\Remote Session Environment
Find the setting: Use hardware graphics adapters for all Remote Desktop Services sessions.
Enable this policy (set to Enabled).
For TurboBoost activation visit the server BIOS using iKVM and go to:
Advanced Setup Configurations → CPU Configuration → Advanced Power Management Configuration
Select
Power Technology = Custom
Became active:
Hardware PM (Power Management) State Control
Enable inner option:
Hardware P-State: native.
Our servers are equipped with Intel® Xeon® Gold 2nd, 3rd, and 5th Generation processors:
The CPU frequency displayed in the guest OS is derived from the processor’s Time Stamp Counter (TSC). On latest CPUs, the TSC runs at a fixed, invariant frequency, which is typically set to either the base (nominal) frequency or the maximum all-core turbo frequency guaranteed under sustained AVX workloads — whichever is lower.
For Intel® Xeon processors, these two values often differ significantly. When AVX2/AVX-512 instructions are used, the CPU dynamically reduces its operating frequency to stay within thermal and power limits. However, the TSC frequency remains constant and is based on the worst-case sustained frequency (e.g., under AVX-512 load), not the peak or current performance frequency.
For example, according to Intel specifications for Cascade Lake and newer families:
Because the TSC runs at this lower, fixed frequency, the value shown in the guest OS does not reflect the actual dynamic clock speed of the CPU cores during operation. The real (effective) frequency changes in real time based on workload, thermal conditions, and instruction set—often running well above the TSC frequency—but the OS still reports the invariant TSC rate.
At immers.cloud, you get access to 13 powerful NVIDIA GPUs. From flagship AI accelerators to high-performance graphic cards for rendering and gaming, we cover it all.
No need to buy or maintain expensive hardware. Run your projects in the cloud with full performance on demand.
For large-scale neural network training and inference, choose our latest models: Tesla A100, H100, and the new H200. These are the foundation of major generative AI, scientific research, and data science projects.
With massive memory capacity, high bandwidth, and next-gen tensor cores, they deliver record-breaking performance. They also handle long-context workloads with ease.
Need rendering, 3D visualization, or gaming? Go with RTX series GPUs: from the proven RTX 2080 Ti to the top-tier RTX 5090.
They excel in Blender, Cinema 4D, Unreal Engine, Redshift, Octane, and other render engines. All support NVENC hardware encoding: perfect for streaming and 4K+ video processing.
For lighter tasks like voice, image, and video processing, Tesla T4, A10, A2, and L4 are ideal.
These are efficient, cost-effective, and perfect for low-latency GPU virtual desktops and services. Stable, reliable, and affordable.
All powered by the cloud. No baremetal. Just results.
A Key Pair is a cryptographic key used to securely connect to Virtual Machines running Linux, as well as to decrypt the password for Virtual Machines running Windows Server.
Important: All Linux distributions (except cirrOS, which does not support ed25519) support public keys in the following formats:
- ecdsa-sha2-nistp256
- ecdsa-sha2-nistp384
- ecdsa-sha2-nistp521
- ed25519
- rsa
Windows Server images support only rsa-formatted public keys, and only if they were generated using the command ssh-keygen -t rsa -m pem or created through the immers.cloud web interface.
If the public key is in an unsupported format, the Key Pair will still be created, but you will not be able to decrypt the password using the private key.
You can also create a Key Pair automatically when creating a Virtual Machine, in the Key Pair field, select Create new Key Pair.
We currently offer over 100 pre-configured images with the following operating systems:
You can also select a pre-configured image from the Marketplace.
Yes, you can upload your own image to the immers.cloud platform: either from a local file or via a direct URL from an external source.
To do this, go to Storage → Images in the left-hand menu and click Create.
In the window that appears, specify:
Once you click Create, the image upload will begin.
Note: If you provide a URL, the image will be downloaded at the moment you create a virtual server.
Important: When creating a server from an ISO image, the entire contents of the ISO are written directly to the server’s primary disk, which can lead to unexpected system behavior.
We recommend:
This approach lets you install an OS from the ISO just like using a physical installation disc—ensuring a clean and predictable setup.
For your image to work correctly on our platform, the following conditions must be met:
Linux Images
Windows Images
If these requirements aren’t met, you may need to manually configure users, drivers, and access after launching the server.
Install NVIDIA drivers from the official archive:
wget http://us.download.nvidia.com/XFree86/Linux-x86_64/440.64/NVIDIA-Linux-x86_64-440.64.run
chmod +x NVIDIA-Linux-x86_64-440.64.run
sudo ./NVIDIA-Linux-x86_64-440.64.run
or use images with the drivers installed: Ubuntu 16.04.6 or Ubuntu 18.04.4 with CUDA.
The Internet connection speed for virtual machines is up to 20 Gbps. You can order extended or guaranteed bandwidth if needed via technical support.
Currently, each server is equipped with two 10GbE network interfaces. This provides high bandwidth, fault tolerance, and flexibility for configuring advanced network scenarios: from load balancing to traffic isolation.
Done!
Now, when creating a virtual machine, you’ll be able to attach it to your private network.
This network is fully isolated from the external environment, only your virtual servers can reside in it.
However, you can assign a floating IP address to any virtual machines in this network to grant it internet access.
Video tutorial
The wlanapi.dll file is part of the Wireless LAN Service component, which is often disabled by default in server editions of Windows (e.g., Windows Server).
To resolve the error:
After the restart, the wlanapi.dll file will be available in the system, and the error will be resolved.
Note: Even on physical servers without a Wi-Fi adapter, this component can still be installed. Some applications require it—even when wireless connectivity is not in use.
If you need to transfer a small file, you can use the clipboard during an RDP session. Copy the file to the clipboard (Ctrl+C) on your local computer or the Virtual Machine, then paste it (Ctrl+V) in the desired location.
Important: Avoid performing other clipboard operations during this process, as they may interfere with the transfer.
To upload or download large volumes of data from a Virtual Machine running Windows Server, you can set up an FTP server.
1. Download and install FileZilla Server using default settings;
2. After installation, the FileZilla Server Interface will launch. In the connection setup window, keep all settings at their defaults;
3. To automatically connect to the server next time, check the box: Always connect to this server;
4. Go to Edit → Users;
5. On the General tab, click Add to create a new user. Enable the Password option and set a password;
6. Switch to the Shared folders tab. Add a folder that the new user will access upon connecting, and define access permissions. You can either create a new folder (e.g., C:\FTP) or select an existing one on the Virtual Machine’s disk.
To allow connections to the FTP server, configure the Windows Firewall on the Virtual Machine:
1. Open the Start menu and select Settings;
2. Search for Firewall and choose Allow an app through Windows Firewall;
3. Click Change settings, then Allow another app;
4. Browse and select the FileZilla Server executable (by default: C:\Program Files (x86)\FileZilla Server\FileZilla Server);
5. After adding it, check both Private and Public boxes.
Use any FTP client (e.g., FileZilla) to connect:
Once connected, you can begin transferring files.
In earlier versions of Windows Server, antivirus settings were only accessible via PowerShell. Now, you can also manage them through the graphical interface: Windows Settings → Windows Security → Virus & threat protection.
If you prefer the command line, open PowerShell as Administrator and run:
Set-MpPreference -DisableRealtimeMonitoring $true.
To re-enable virus and threat protection, run:
Set-MpPreference -DisableRealtimeMonitoring $false.
Both methods—GUI and command line—are fully functional.
When attempting to change the time zone, some users may encounter an error:
Operation could not be completed. You do not have permission to perform this task. Contact your network administrator.
There are two ways to resolve this issue:
First metjod. Change time zone via Control Panel as Administrator. Click Start, type timedate.cpl in the search bar, and press Enter. Right-click on the Date and Time window and select Run as administrator;. Adjust the time zone settings as needed and save the changes.
Second method. Grant user rights via Group Policy^
gpedit.msc, and launch the Group Policy Editor;In some cases Windows Task Manager displays incorrect GPU Load information. Try to use any other GPU monitoring tools, please.
cutil64.dll was a part of CUDA library version 4.2 and earlier. Find this file in CUDA SDK (for example, version 4.2) and copy it to an application folder.
To install the .NET Framework 3.5, use PowerShell launched as an administrator.
After opening PowerShell with administrative privileges, run the following command:
Install-WindowsFeature NET-Framework-Core
Verify that .NET 3.5 has been installed by running the command again and confirming that the status shows Success: True.
Run the installer as an administrator.
Steps to add a path to the PATH environment variable:
The new path is now added to the system PATH variable and will be available for use in Command Prompt or PowerShell after restarting the session.
The minimum payment amount is 25 EUR. To pay with a card issued outside the Russian Federation (and which is not a part of the UnionPay payment system), please contact our managers via Telegram or sale@immers.cloud
We offer the pay-as-you-go billing, and your balance updates every hour.
immers.cloud provides access to modern GPU accelerators with flexible payment. You pay only for the actual runtime of your server. VM billing starts the moment the instance reaches the Active status. This solution works for both individual users and teams of any budget.
To avoid paying for an idle virtual machine, use the Shelve feature. It archives your server, releases compute resources, and stops hardware billing. After shelving, you only pay for snapshot (or persistent volume) storage and, if needed, a reserved IP address (learn more about this feature here).
To set up notifications, go to your profile, and click the envelope icon next to your name. Then select the balance threshold at which you want to receive reminders, and choose your preferred notification method — text-messages or email.
Important: To receive SMS-notifications, you must first verify your mobile phone number.
You can pay via Stripe using the payment link. The minimum amount is 25 EUR.
After completing the payment, please send the following information in the support chat:
Important notes:
Please note: Our team operates Monday to Friday, 10:00 a.m. to 6:00 p.m. Moscow time (GMT+3). We do not process payments on weekends (Saturday and Sunday).
If you need assistance, our support team is available via:
Customer Support Team:
GPU cloud servers are available with long-term plans ranging from 1 to 12 months.
To use this option, select a prepaid plan when creating a server. You can also switch an already running server from pay-as-you-go billing to a long-term plan. In both cases, the full amount for the entire period will be charged upfront, so your account balance must have sufficient funds before creating the cloud server.
Important: Once switched, you cannot revert to per-second billing.
You also won’t be able to delete the server before the hosting period expires.
Current available plans are listed on the Pricing page.
Virtual Machines are billed in all statuses except Build and Error.
Important: Please read why charges apply to powered-off Virtual Machines.
Billing for all paid resources is calculated on a pay-as-you-go basis, but your account balance is updated hourly in the user portal. Detailed charge breakdowns for all paid items are available in the Purchases section.
The cost of a Configuration is calculated according to the current Pricing table and equals the sum of the following components:
The total cost of a Virtual Machine is typically higher than the base Configuration cost.
Additional charges apply for:
Charges for powered-off Virtual Machines apply only to reserved resources, which remain allocated even when the machine is stopped. These include:
Please note that vCPU and RAM are not billed while the Virtual Machine is powered off.
Important: To avoid charges for inactive machines, learn how to stop paying for a Virtual Machine that is currently not in use.
To download files from a Virtual Machine running Linux, set up an SFTP connection using FileZilla.
Download the private key file with the .PPK extension from the Key Pairs section.
Configure the connection to your Virtual Machine in FileZilla:
Video tutorial
Update your operating system by running the following commands in sequence:
sudo apt update
sudo apt upgrade
Set a password for your user account: sudo passwd ubuntu
Install the Xfce desktop environment and the xrdp remote desktop server: sudo apt install xfce4 xrdp
Allow X server startup for any user: sudo nano /etc/X11/Xwrapper.config =anybody
Note: If you prefer to edit the file manually:
sudo nano /etc/X11/Xwrapper.configallowed_users=anybodySet Xfce as your default desktop session: echo xfce4-session >~/.xsession
Restart xrdp to apply the changes: sudo systemctl restart xrdp
Verify that xrdp is running: sudo systemctl status xrdp
If your virtual machine has a GPU, run these extra commands for better compatibility:
sudo usermod -aG ssl-cert xrdp
sudo apt install xorgxrdp
sudo systemctl restart xrdp
Use the Windows Remote Desktop Connection client:
ubuntu) and the password you set earlier/Important: Hardware GPU acceleration is not available over xrdp. Some graphics-intensive applications may not function correctly or may run in software rendering mode.
To restore your server’s operation, we provide a Rescue mode for cloud servers, accessible via the Actions dropdown menu.
Rescue mode boots your server from a clean operating system image while preserving all data on your original disk. Once booted into Rescue mode, you can connect to the server and perform recovery operations.
Note: To access your data, you must first mount your server’s disk: sudo mount /dev/vda1 /mnt
Optionally, you can use chroot to work inside your original operating system environment:
sudo mount -t proc proc /mnt/proc
sudo mount -t sysfs sys /mnt/sys
sudo mount -o bind /dev /mnt/dev
sudo mount -t devpts pts /mnt/dev/pts
chroot /mnt /bin/bash -i
From here, you can make any necessary changes to restore access. For example, disabling a firewall, fixing configuration files, or troubleshooting other issues.
Once recovery is complete, return your server to normal operation by selecting Unrescue from the Actions menu.
If you lost your Key Pair, did not set a password for emergency access via VNC, or misconfigured the network and cannot connect to your CentOS Linux Virtual Machine (VM), this guide will help you restore access.
You need to perform the following steps to regain SSH access to your CentOS-based VM:
sudo systemctl status sshd or sudo journalctl -xesudo ssh-keygen -t rsa -f /etc/ssh/ssh_host_rsa_key
sudo ssh-keygen -t ecdsa -f /etc/ssh/ssh_host_ecdsa_key
sudo ssh-keygen -t ed25519 -f /etc/ssh/ssh_host_ed25519_keysudo systemctl start sshd. You can verify that the SSH server is running: sudo systemctl status sshdsudo vi /etc/sysconfig/selinux and change enforcing to disabled, then save the changes.sudo getenforceA Volume is a virtual hard disk located on distributed network storage.
When creating Volume-backed Virtual Machines, a Volume is created automatically and used as the boot disk. A Volume containing a configured operating system and user data can be reused to create a new Virtual Machine. For more details, see: How to avoid paying for a Virtual Machine that is currently not in use.
Important: If you do not plan to use the automatically created Volume in the future, you must delete it manually after removing the Virtual Machine.
You can also create Volumes of any size and attach them to Virtual Machines as additional disks.
You can store files in S3 cloud storage. To get started, you will need additional software — an S3 client.
To begin using the storage, create an S3 user or access credentials in your user account.
Creating an S3 user and Connection Credentials:
Navigate to S3 User Management to create a new user.
Specify the S3 ID and User Name (you can choose any name).
Once the user is created, the Access Key and Secret Key will be generated and displayed.
Go to Login Credentials to generate login details.
All fields will be filled in automatically; simply save the credentials.
The Access and Secret keys will also be available there.
Use any S3 client to connect to S3 storage, such as S3 Browser.
Setting Up S3 Browser
Follow these steps to configure the S3 Browser:
Open the Accounts menu and select Add New Account.
Fill out the fields on the "New User" page:
Display Name: Enter any name (this is for local display only).
Account Type: Choose "S3 Compatible Storage."
REST Endpoint: Enter https://api.immers.cloud:8080.
Access Key ID: Use the Access Key from the S3 user details.
Secret Access Key: Use the Secret Key from the S3 user details.
Advanced Settings:
Set the Addressing Model to Path Style.
Save the configuration.
Click Next, then Add New Account.
Your account is now ready. You can create buckets (referred to as "baskets" in the interface) and upload files.
Managing Buckets
Buckets are displayed on the website in the Buckets Section.
You can monitor your bucket activity in the Bucket Operations Section.
You can transfer Volumes between accounts. To do this:
Important: If the sender account has a negative balance, you must top it up to at least zero before transferring. The recipient account must also have a positive balance.
Go to the Connections section and click Create. Then select the Virtual Machine and the Volume you want to attach. Choose the access mode:
ro — read only;
rw — read write (read and write).
Click Attach.
A Backup is typically used to preserve the data or state of a single Virtual Machine on a regular, recurring schedule.
Use a Snapshot when you plan to create multiple Virtual Machines from the same source image.
A Snapshot (also called a snap) is a one-time image of a Virtual Machine’s boot disk, manually created by the user.
A Backup is also an image of the boot disk, but it is taken on a regular schedule (daily, weekly, monthly), with the option to set the number of backup rotations retained for a single Virtual Machine.
After creating the Volume attachment, ensure the Volume status has changed to "In Use". Then, use the built-in application Computer Management, go to Storage → Disk Management, to initialize the new disk, assign a drive letter, and make it available.
Shut down the Virtual Machine using the Stop command in the Actions dropdown menu. Wait until the status changes to Shutoff.
In the Actions dropdown menu, select Create Snapshot, enter a name for the snapshot, and click Create Snapshot. Wait until the status becomes Active.
You can now select this Snapshot when creating a new Virtual Machine.
Note that Snapshots can only be created for Local-type Virtual Machines.
To expand a Volume, click the icon. Enter the new desired size and click Expand. Wait for the status to change to Available.
If the Volume is currently attached to a Virtual Machine, restart the Virtual Machine to see the changes in the guest operating system.
Important: The Volume size can only be increased, not decreased.
To create a Virtual Machine (VM) with a local boot disk (Local), specify the `"imageRef"` parameter in the API request.
In the example below, a VM is created with a new local boot disk from the pre-installed image 754e77f5-9a0c-4549-b49d-7bde34dc960f (Ubuntu 18.04.4 CUDA).
{
"server" : {
"name": "local",
"imageRef": "754e77f5-9a0c-4549-b49d-7bde34dc960f",
"flavorRef": "8f9a148d-b258-42f7-bcc2-32581d86e1f1",
"availability_zone": "nova",
"networks": [{ "uuid": "cc5f6f4a-2c44-44a4-af9a-f8534e34d2b7" }],
"key_name": "my_keypair",
"max_count": 1,
"min_count": 1
}
}
Where:
"name": "local" — the name of the new VM;"imageRef": "754e77f5-9a0c-4549-b49d-7bde34dc960f" — OpenStack ID of the OS image;"flavorRef": "8f9a148d-b258-42f7-bcc2-32581d86e1f1" — OpenStack ID of the configuration;"networks": [{ "uuid": "cc5f6f4a-2c44-44a4-af9a-f8534e34d2b7" }] — OpenStack ID of the network in which the VM will receive an IP-address;"key_name": "my_keypair" — the name of your Key Pair.When this Virtual Machine is deleted, its boot disk is also permanently removed.
To create a Volume-backed Virtual Machine (with a boot disk on network block storage), include the block_device_mapping section in your API request. You can create boot volumes from an Image "source_type": "image", a Volume Snapshot "source_type": "snapshot", or attach an existing Volume "source_type": "volume".
In the example below, a new Virtual Machine is created with a boot Volume built from image 754e77f5-9a0c-4549-b49d-7bde34dc960f (Ubuntu 18.04.4 CUDA).
{
"server" : {
"name": "volume-backed",
"block_device_mapping_v2": [
{
"uuid": "754e77f5-9a0c-4549-b49d-7bde34dc960f",
"source_type": "image",
"destination_type": "volume",
"boot_index": 0,
"volume_size": "20",
"delete_on_termination": false
}
],
"flavorRef": "8f9a148d-b258-42f7-bcc2-32581d86e1f1",
"availability_zone": "nova",
"networks": [{ "uuid": "cc5f6f4a-2c44-44a4-af9a-f8534e34d2b7" }],
"key_name": "my_keypair",
"max_count": 1,
"min_count": 1
}
}
Where:
"name": "volume-backed" — name of the new Virtual Machine;"uuid": "754e77f5-9a0c-4549-b49d-7bde34dc960f" — OpenStack ID of the source image;"source_type": "image" — the boot Volume will be created from an image;"volume_size": "20" — size of the new Volume in gigabytes;"delete_on_termination": false — the boot Volume will not be deleted when the VM is deleted;"flavorRef": "8f9a148d-b258-42f7-bcc2-32581d86e1f1" — OpenStack ID of the configuration;"networks": [{ "uuid": "cc5f6f4a-2c44-44a4-af9a-f8534e34d2b7" }] — OpenStack ID of the network where the VM will receive an IP address;"key_name": "my_keypair" — name of your Key Pair.If delete_on_termination is set to true the boot Volume will be deleted when the Virtual Machine is removed. When creating a VM from a Volume Snapshot, the Snapshot and its parent Volume are preserved regardless of this setting.
Setting delete_on_termination to false, ensures the boot Volume persists after VM deletion, allowing you to reuse it later.
In the working directory of the Moonlight.exe in the latest Moonlight-xxxxxxxxxx.log you can find:
00:00:01 - SDL Info (0): GPU driver: igdumdim64.dll 20.19.15.4377
00:00:01 - SDL Info (0): Detected buggy Intel GPU driver installed. Update your Intel GPU driver to enable HEVC!
To solve this issue install a newer driver from Intel: https://www.intel.ru/content/www/ru/ru/download/19344/intel-graphics-windows-dch-drivers.html
Launch the Windows PowerShell application.
Enter the following command: (gwmi Win32_Bus -Filter 'DeviceID like "PCI%"').GetRelated('Win32_PnPEntity').GetDeviceProperties('DEVPKEY_Device_Address').deviceProperties | ft DeviceID,keyName,{$_.data -shr 16},{$_.data -band 0xFFFF} then press Enter.
Run the command: lspci -d 10de:. Where 10de is the vendor ID for NVIDIA.
The VNC console is available in the Actions dropdown menu and runs directly in the browser window.
The connection is established to a VNC server on the hypervisor, not on the individual Virtual Machine. It cannot be accessed using an external VNC client. This tool is intended for emergency monitoring only, not for regular use.
Important: For Virtual Machines with a GPU that includes a display emulator in their configuration, the VNC console on the website will not function.
Users who need full remote access via VNC must set up their own VNC server within the guest operating system using a third-party VNC client.
GRID is a specialized NVIDIA driver that enables Tesla GPUs. Originally designed only for compute workloads to support graphics-intensive tasks such as:
Important: The GRID driver works only with Tesla GPUs. It does not support RTX graphic cards, as they natively operate in graphics mode using standard drivers.
By default, when using GRID on our platform, GPUs operate in WDDM mode.
To switch GPU mode, run the following command as administrator:
nvidia-smi -g {GPU_ID} -dm {0|1}
Note: Mode changes require a server reboot to take effect.
When installing the GRID driver manually, configure the license server as follows:
Our license server is stable and available to all immers.cloud customers using NVIDIA GRID.
Building an immersion data center is significantly cheaper than an air-cooled one.
Immersion cooling is currently the most cost-effective way to cool high‑density IT equipment. Unlike a traditional air‑cooled data center, our system uses only 5% of the total energy for cooling.
Our IT hardware operates without fans, reducing power consumption by up to 30% right out of the box. Each of our 42U racks can accommodate up to 160 GPUs with a power load of up to 80 kW per rack.
Immersion cooling keeps components at optimal temperatures, ensuring stability and enabling higher clock speeds for every CPU and GPU.
Immersion cooling is a method of cooling heat-generating equipment by submerging it in liquid. In the case of the immers.cloud cloud platform, we submerge all servers and networking equipment into a special dielectric fluid that removes heat from server components 1,200 times more efficiently than air.
An immersion data center is significantly smaller and simpler to maintain than an air-cooled data center. It is not only safe, but also environmentally friendly. We built a 500 kW data center, while our competitors are forced to build equivalent air-cooled facilities of 1.5 MW. While our competitors release heat into the atmosphere, we reuse the recovered heat for hot water supply (in summer) and heating (in winter). We do not use refrigerants that harm our planet’s atmosphere.