HPC - User contributions [en-gb]

DAIM-Guide/Python

2025-10-03T09:16:55Z

Chris.collins:

== Installing Packages ==
<pre>
!conda install -y python_package
</pre>

== Checking for GPU Access ==
Even if you have requested a Jupyter session on one of the GPU resources, it is important to check that you actually have GPU access for your notebook when first setting up your workflow. You can do this as follows:

With Torch:
<pre>
>>> import torch
>>> torch.cuda.is_available()
True
</pre>
If this reports 'False' then Torch cannot find a valid GPU device. Similarly, with Tensorflow you can check with the following. If you don't see a similar output referencing a device, and instead see an error mentioning not being able to load libcuda library or indicating nvidia drivers not existing, then the Tensorflow cannot find the GPU device:
<pre>
>>> import tensorflow as tf
>>> tf.config.list_physical_devices('GPU')
[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]
</pre>

If your session does not find a GPU device, please confirm you are running on a GPU enabled system, the following should report a gpu hostname (e.g. gpu01 - gpu09) etc:

<pre>
!hostname
</pre>

Please report any issues via the Support portal (see [[General/GettingHelp| Getting help]])

General/GettingHelp

2025-08-01T06:54:17Z

Chris.collins:

Please raise all requests for support via the University support hub for staff and Hubble for students, providing as much information as possible to help us help you - for example when reporting issues please ensure you include the following information:

* Full details about the issue you are seeing, what you need to know and the software you want assistance with etc.
* Job ID.
* Session ID (if running on Viper OnDemand).
* The names and paths of any relevant files, including job scripts, Jupyter Notebooks, and Virtual Environments etc.
* Any further background information that you think may be useful (has this task worked previously, what has changed, what you have tried etc).

The following links will take you to the appropriate forms:
{| class="wikitable"
|- style="vertical-align:top;
! Support Request !! Student Links !! Staff Links
|- style="vertical-align:top;
| style="width: 30%" | '''Request Viper Support''' || style="width: 35%" | [https://hull.service-now.com/student?id=sc_cat_item&table=sc_cat_item&sys_id=51fcd7a093b8c21001517524fbba1014 Hubble link] || style="width: 35%" | [https://hull.service-now.com/staff?id=sc_cat_item&table=sc_cat_item&sys_id=cb1856289334c21001517524fbba1046 Staff Support Hub link]
|- style="vertical-align:top;
| style="width: 30%" | '''Request Job Extension''' || style="width: 35%" | [https://hull.service-now.com/student?id=sc_cat_item&table=sc_cat_item&sys_id=6cbfbd351ba182d07422eac8b04bcb2b Hubble link] || style="width: 35%" | [https://hull.service-now.com/staff?id=sc_cat_item&table=sc_cat_item&sys_id=256d39fd1b6182d07422eac8b04bcb3c Staff Support Hub link]
|- style="vertical-align:top;
| '''Viper Account Application''' || [https://hull.service-now.com/student?id=sc_cat_item&table=sc_cat_item&sys_id=6f290b751b2982d07422eac8b04bcbf1 Hubble link] Note: For DAIM MSc students, Viper access is requested by your project supervisor. Please discuss requirements with them and they can request an account with the corresponding link Note: For Undergraduate students, please user the Hubble link above and provide details of your project supervisor in the form and we will get confirmation from them regarding your proposed use.|| [https://hull.service-now.com/staff?id=sc_cat_item&table=sc_cat_item&sys_id=15a44bbd1be582d07422eac8b04bcb5f Staff Support Hub link] [https://hull.service-now.com/staff?id=sc_cat_item&table=sc_cat_item&sys_id=5f7e08cb1b694610998b6204b24bcbbf Supervisor link for DAIM MSc Accounts]
|- style="vertical-align:top;
| '''Request VPN Access''' (this is not managed by the HPC team) || [https://hull.service-now.com/student?id=sc_cat_item&table=sc_cat_item&sys_id=42bcf9bd1b6182d07422eac8b04bcb93 Hubble link] || [https://hull.service-now.com/staff?id=sc_cat_item&table=sc_cat_item&sys_id=6c2939b91b6182d07422eac8b04bcbef Staff Support Hub link]
|-
|}

2025-07-02T14:42:21Z

Chris.collins: /* Checking for GPU Access */

= DAIM

== Installing Packages ==
<pre>
!conda install -y python_package
</pre>

== Checking for GPU Access ==
Even if you have requested a Jupyter session on one of the GPU resources, it is important to check that you actually have GPU access for your notebook when first setting up your workflow. You can do this as follows:

With Torch:
<pre>
>>> import torch
>>> torch.cuda.is_available()
True
</pre>
If this reports 'False' then Torch cannot find a valid GPU device. Similarly, with Tensorflow you can check with the following. If you don't see a similar output referencing a device, and instead see an error mentioning not being able to load libcuda library or indicating nvidia drivers not existing, then the Tensorflow cannot find the GPU device:
<pre>
>>> import tensorflow as tf
>>> tf.config.list_physical_devices('GPU')
[PhysicalDevice(name='/physical_device:GPU:0', device_type='GPU')]
</pre>

If your session does not find a GPU device, please confirm you are running on a GPU enabled system, the following should report a gpu hostname (e.g. gpu01 - gpu09) etc:

<pre>
!hostname
</pre>

Please report any issues via the Support portal (see [[General/GettingHelp| Getting help]])

DAIM-Guide/Python

2025-07-02T14:40:11Z

Chris.collins: /* Checking for GPU Access */

2024-10-17T21:00:46Z

Chris.collins: Chris.collins uploaded a new version of File:Viperondemand.png

DAIM Guide

2024-10-07T08:48:05Z

Chris.collins: /* Prerequisites */

__TOC__

== Introduction ==
Viper, the University of Hull’s High-Performance Computing facility, is used by research staff and students in many disciplines across the University and is a potentially significant tool for those with any computational element to their research.
Viper is a ‘cluster’ of approximately 200 computes built into a cluster and features more than 6000 compute cores, high memory systems and GPU technology together with dedicated high-performance storage and fast interconnect between systems to meet the needs of the most computationally demanding tasks.
Some notes about
* Viper runs Linux, though to make use of Viper you do not need to be a Linux expert – being familiar with just a few commands is enough to get started.
* Viper runs a scheduler, a piece of software which manages access to the computers across the cluster, monitoring what is running and how resources such as memory, CPU processors and GPU cards are being used.
* Users run ‘jobs’ some of which are interactive (like Jupyter Notebook sessions) but many of which are submitted to run automatically without user interaction. Jobs usually have a batch script, a text file containing the recipe for the job, for example, what resource is needed and what the job should do.

== Prerequisites ==
* Viper account - '''Access for a Viper account should be completed by the MSc project supervisor on behalf of the student and confirms that the requirements for Viper access have been discussed and are supported by the project supervisor.''' To apply for a Viper account, please see [https://hull.service-now.com/staff?id=sc_cat_item&table=sc_cat_item&sys_id=5f7e08cb1b694610998b6204b24bcbbf DAIM MSc Project Account Request]
* VPN - You will need a VPN account to access Viper from off-campus. Please see [https://hull.service-now.com/ GlobalProtect Client VPN Service page on the Support Portal]

== Available Resource ==
'''Important''' While Viper has a range of GPU resources available for DAIM student use, including Nvidia A100 and Nvidia A40-based systems, such resource is in very high demand and as such at times there may be many people competing for GPU resources meaning long pend (queuing) times are likely.

Viper is a shared resource, so to help alleviate this for all users, could we suggest:

* Where possible, try and only run '''final''' GPU runs on Viper, and '''keep development of code to either other systems (e.g. own computer, DAIM lab computer or Viper CPU nodes)'''
* Don't request longer runtimes than are required or if you have resources allocated to you but are not going to work for a period of time ''' please cancel your allocation'''
* Please consider other users require GPU access, '''don't queue up multiple Jupyter sessions'''

{| class="wikitable" style="text-align:center;"
|- style="text-align:left;"
! Resource Type
! GPU RAM *
! System RAM
! Count
! Accessible as
! Notes
|-
| style="text-align:left;" | Nvidia A100 MIG 40GB
| 40GB
| 64GB
| 4
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | High demand
|-
| style="text-align:left;" | Nvidia A100 MIG 20GB
| 20GB
| 64GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | High demand
|-
| style="text-align:left;" | Nvidia A100 MIG 10GB
| 10GB
| 32GB
| 1
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | Medium demand

|-
| style="text-align:left;" | CPU only
| -
| 128GB / 1TB
| 10+
| style="text-align:left;" | [[DAIM_Guide#Pilot_Web_Portal_Access Ondemand | Web portal]]
| style="text-align:left;" | Lower demand but low performance for ML. Use for development and testing

|-
| style="text-align:left;" | NVidia A40
| 48GB
| 128GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A40_GPU_access | Web portal]]
| style="text-align:left;" | Very high demand, expect long wait times - these should only be used if the task exceeds GPU memory available on MIG instances
|}

<nowiki>*</nowiki> There are two specifications of DAIM lab PCs, the majority are equipped with Nvidia GeForce RTX 3070 GPU cards, with 8GB of GPU memory - all GPU instances on Viper have at least an equivalent amount of GPU memory available. There are also several lab PCs (dual monitor lab machines) with Nvidia RTX 3090 GPU cards with 24GB of GPU memory.

== OnDemand Web Portal Access (pilot) ==
We are currently testing a web route to provide Jupyter access via our pilot Viper OnDemand web portal - see http://hpc.mediawiki.hull.ac.uk/General/OOD

This service is in testing, please raise any issues via the University support portal, searching for '''Viper''' to find forms to help you request support.

'''Please note:'''
* VPN access is still required to connect to the Viper OnDemand web portal.
* The DAIM Jupyter apps are in testing and there may be issues which have yet to be resolved.

If you connect to the web portal, you will see several DAIM prefixed apps, which can be used to connect:
==== Create Python Virtual Environment ====
* '''DAIM Create Environment''' - this will create a base Python virtual environment. You can change the name of the base environment by changing the text in the Environment Name box. You can then select this environment kernel to use when in DAIM session notebooks.

==== Web Portal Nvidia A40 GPU access ====
* '''DAIM Jupyter''' - this will launch a Jupyter lab session in your browser tab
* '''DAIM Jupyter Desktop''' - similarly, this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) access ====
* '''DAIM Jupyter Desktop MIG''' - this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.
* '''DAIM Jupyter MIG''' - this will launch a Jupyter lab session in your browser tab. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) automated run ====
* '''TEST DAIM nbconvert Job''' - This will submit a task to run Jupyter-nbconvert to automatically execute the specific notebook when the resource is available without you needing to interact with the notebook. Specify a notebook using the "select path" button and click launch. Once complete nbconvert will produce an output notebook with nbconvert in the name (e.g. the input mynotebook.ipynb will produce mynotebook.nbconvert.ipynb). After you launch the task, it will just run automatically when the resource is available without you needing to interact with it and will be listed as "Running" in your session list. Note that there is very little in terms of output until the notebook execution is complete. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal GPU and Job Status Page ====
To help work out what GPU resources are available for use, and how long a queued job may take, there is a monitor tool on the web portal accessible via '''Jobs > GPU Status and current jobs'''.

'''Warning:''' The first time this is run you will be presented with a warning about needing to restart your web server - doing this may terminate running Jupyter sessions, so do not proceed if you have active sessions.

DAIM Guide

2024-10-07T08:46:11Z

Chris.collins:

__TOC__

== Introduction ==
Viper, the University of Hull’s High-Performance Computing facility, is used by research staff and students in many disciplines across the University and is a potentially significant tool for those with any computational element to their research.
Viper is a ‘cluster’ of approximately 200 computes built into a cluster and features more than 6000 compute cores, high memory systems and GPU technology together with dedicated high-performance storage and fast interconnect between systems to meet the needs of the most computationally demanding tasks.
Some notes about
* Viper runs Linux, though to make use of Viper you do not need to be a Linux expert – being familiar with just a few commands is enough to get started.
* Viper runs a scheduler, a piece of software which manages access to the computers across the cluster, monitoring what is running and how resources such as memory, CPU processors and GPU cards are being used.
* Users run ‘jobs’ some of which are interactive (like Jupyter Notebook sessions) but many of which are submitted to run automatically without user interaction. Jobs usually have a batch script, a text file containing the recipe for the job, for example, what resource is needed and what the job should do.

== Prerequisites ==
* Viper account - to apply for a Viper account, please see [https://hull.service-now.com/staff?id=sc_cat_item&table=sc_cat_item&sys_id=5f7e08cb1b694610998b6204b24bcbbf DAIM MSc Project Account Request]
* VPN - You will need a VPN account to access Viper from off-campus. Please see [https://hull.service-now.com/ GlobalProtect Client VPN Service page on the Support Portal]

== Available Resource ==
'''Important''' While Viper has a range of GPU resources available for DAIM student use, including Nvidia A100 and Nvidia A40-based systems, such resource is in very high demand and as such at times there may be many people competing for GPU resources meaning long pend (queuing) times are likely.

Viper is a shared resource, so to help alleviate this for all users, could we suggest:

* Where possible, try and only run '''final''' GPU runs on Viper, and '''keep development of code to either other systems (e.g. own computer, DAIM lab computer or Viper CPU nodes)'''
* Don't request longer runtimes than are required or if you have resources allocated to you but are not going to work for a period of time ''' please cancel your allocation'''
* Please consider other users require GPU access, '''don't queue up multiple Jupyter sessions'''

{| class="wikitable" style="text-align:center;"
|- style="text-align:left;"
! Resource Type
! GPU RAM *
! System RAM
! Count
! Accessible as
! Notes
|-
| style="text-align:left;" | Nvidia A100 MIG 40GB
| 40GB
| 64GB
| 4
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | High demand
|-
| style="text-align:left;" | Nvidia A100 MIG 20GB
| 20GB
| 64GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | High demand
|-
| style="text-align:left;" | Nvidia A100 MIG 10GB
| 10GB
| 32GB
| 1
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | Medium demand

|-
| style="text-align:left;" | CPU only
| -
| 128GB / 1TB
| 10+
| style="text-align:left;" | [[DAIM_Guide#Pilot_Web_Portal_Access Ondemand | Web portal]]
| style="text-align:left;" | Lower demand but low performance for ML. Use for development and testing

|-
| style="text-align:left;" | NVidia A40
| 48GB
| 128GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A40_GPU_access | Web portal]]
| style="text-align:left;" | Very high demand, expect long wait times - these should only be used if the task exceeds GPU memory available on MIG instances
|}

<nowiki>*</nowiki> There are two specifications of DAIM lab PCs, the majority are equipped with Nvidia GeForce RTX 3070 GPU cards, with 8GB of GPU memory - all GPU instances on Viper have at least an equivalent amount of GPU memory available. There are also several lab PCs (dual monitor lab machines) with Nvidia RTX 3090 GPU cards with 24GB of GPU memory.

== OnDemand Web Portal Access (pilot) ==
We are currently testing a web route to provide Jupyter access via our pilot Viper OnDemand web portal - see http://hpc.mediawiki.hull.ac.uk/General/OOD

This service is in testing, please raise any issues via the University support portal, searching for '''Viper''' to find forms to help you request support.

'''Please note:'''
* VPN access is still required to connect to the Viper OnDemand web portal.
* The DAIM Jupyter apps are in testing and there may be issues which have yet to be resolved.

If you connect to the web portal, you will see several DAIM prefixed apps, which can be used to connect:
==== Create Python Virtual Environment ====
* '''DAIM Create Environment''' - this will create a base Python virtual environment. You can change the name of the base environment by changing the text in the Environment Name box. You can then select this environment kernel to use when in DAIM session notebooks.

==== Web Portal Nvidia A40 GPU access ====
* '''DAIM Jupyter''' - this will launch a Jupyter lab session in your browser tab
* '''DAIM Jupyter Desktop''' - similarly, this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) access ====
* '''DAIM Jupyter Desktop MIG''' - this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.
* '''DAIM Jupyter MIG''' - this will launch a Jupyter lab session in your browser tab. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) automated run ====
* '''TEST DAIM nbconvert Job''' - This will submit a task to run Jupyter-nbconvert to automatically execute the specific notebook when the resource is available without you needing to interact with the notebook. Specify a notebook using the "select path" button and click launch. Once complete nbconvert will produce an output notebook with nbconvert in the name (e.g. the input mynotebook.ipynb will produce mynotebook.nbconvert.ipynb). After you launch the task, it will just run automatically when the resource is available without you needing to interact with it and will be listed as "Running" in your session list. Note that there is very little in terms of output until the notebook execution is complete. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal GPU and Job Status Page ====
To help work out what GPU resources are available for use, and how long a queued job may take, there is a monitor tool on the web portal accessible via '''Jobs > GPU Status and current jobs'''.

'''Warning:''' The first time this is run you will be presented with a warning about needing to restart your web server - doing this may terminate running Jupyter sessions, so do not proceed if you have active sessions.

File:OOD-BadResponse.png

2024-10-05T20:53:15Z

Chris.collins: Viper OnDemand Bad Response Error

Viper OnDemand Bad Response Error

Training/Transferring Data

2024-09-23T12:13:25Z

Chris.collins: /* Linux */

__TOC__

==Methods==

===Small to medium amounts of data===

This can be carried out using a number of SFTP-styled packages such as sftp, scp, rsync, and Filezilla to name a few. Filezilla can be downloaded as a Windows application and may be more suitable for people who prefer a Windows environment, rather than the command line.

'''Note''' FTP (insecure) is disabled on viper.hull.ac.uk and SFTP (secure FTP) is the recommended method.

===Large to a huge amount of data (100GB +)===
To avoid excessive network congestion and the long period to transfer such data please raise a [https://hull.service-now.com/ support ticket] for further details. This may typically involve connecting a USB data drive directly to VIPER to copy the data quickly into your user area.

===OneDrive to Viper===

With the University using MS OneDrive, it is possible to copy files to Viper. For multiple files, zip your files up on OneDrive first (considerably easier), as shown here:

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | From OneDrive to Viper
|-
|
* Starting with the zipped file MyData.zip on OneDrive, copied previously and compressed from local storage.
* Click on the '''share icon''' or options for the file here (three horizontal dots) and select '''share'''.

[[File:onedrive-01.png]]

* Then

[[File:onedrive-02.png]]

* Click on '''Copy'''

[[File:onedrive-03.png]]

* On your Viper console application, type '''wget''' and paste in the copied link and follow with '''-O yourfilename''' (otherwise Linux will name the file to the OneDrive link - ''argh!'')

[[File:onedrive-04.png]]

* And here is the received file

[[File:onedrive-05.png]]

'''Remember''' to un-share the file on OneDrive and use unzip (if required) on Linux to restore the file structure.

|}

==Universal==
The following can only be performed on Windows after installing MobaXterm or PuTTY.

The example user is user1; you will need to enter your username here:

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | * To Viper from local computer
|-
|scp myfile user1@viperlogin.hpc.hull.ac.uk:myfile
|-
|rsync -av my-directory viper.hull.ac.uk:/data/users/YourID
|}

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | * From Viper to your local computer
|-
| scp user1@viperlogin.hpc.hull.ac.uk:bio/YourUCInetID/hpc-file
|-
|rsync -av viper.hull.ac.uk:/data/users/YourID/mydatadirectory
|}

==Windows==
The following example uses MobaXterm, for transferring large files it is recommended to use FileZilla. For instructions on how to use FileZilla view the Linux section. follow the instruction in the Getting Connected section for installing MobaXterm.

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | * Uploading Files
|-
|* To upload files to Viper, navigate to the directory you wish to upload files to use the MobaXterm file browser.
* Choose the file/files to upload and click open or alternatively drag and drop the file/files you wish to upload over the Viper file browser.
* The upload may take some time depending on upload speed and file size.
|}

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | * Downloading Files
|-
|* To download files from Viper, navigate to the directory using the MobaXterm file browser.
[[File:mobaxterm_file_browser.jpg]]
* Select the file/files you wish to download and click the 'download selected files button'.
[[File:mobaxterm_download_icon.jpg]]
* Choose the directory where you want the files to download and click ok.
* The download will start, depending on download speed and file size this may take some time.
|}

==MacOS==
The following example uses Cyberduck, you can download Cyberduck from https://cyberduck.io macOS also supports FileZilla for an example of how to use FileZilla please view the Linux section.

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | Connecting
|-
|
* Click Open Connection.
* From the drop-down menu select '''SFTP''' (SSH File Transfer Protocol)
* Enter server viperlogin.hpc.hull.ac.uk, Port: 22, Username: University UID, Password: university password
* Click Connect.

|}

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | Uploading Files
|-
|
* Click Upload
* Choose the file/folder you want to upload and click Choose.
* The transfer window will appear.
* Please note due to file size the upload may take some time.
|}

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | Downloading Files
|-
|
* From the file browser right-click on the file/folder to download and select download from the menu.
* The transfer window will appear.
* Please note due to file size the download may take some time.
|}

==Linux==

The following example uses FileZilla, you can download FileZilla from https://filezilla-project.org/ (''Windows/Linux and MacOS'').
* '''Note''': the FileZilla installer sometimes comes with additional trial software packages. We would advise against installing these

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | Using FileZilla
|-
|
*Enter the Host: viperlogin.hpc.hull.ac.uk, username, password, and port: 22, using an SFTP connection.
[[File:Filezilla_connection_viperlogin.png]]

* A pop-up will appear informing that the server host key is unknown. Click Always trust this host and click ok. In the remote site box, your Viper directory will appear.
* To transfer files click and drag the file/folder from either your pc to Viper (left to right) or from Viper to your pc (right to left). The transfer will start depending on transfer speed and file size this may take some time.

[[File:Filezilla_transferring_files.jpg]]

|}

==Navigation==

* [[Main_Page|Home]]
* [[Applications|Application support]]
* [[General|General]] *
* [[Programming|Programming support]]

File:Filezilla connection viperlogin.png

2024-09-23T12:10:57Z

Chris.collins:

Training/Transferring Data

2024-09-23T12:08:34Z

Chris.collins: /* Linux */

__TOC__

==Methods==

===Small to medium amounts of data===

This can be carried out using a number of SFTP-styled packages such as sftp, scp, rsync, and Filezilla to name a few. Filezilla can be downloaded as a Windows application and may be more suitable for people who prefer a Windows environment, rather than the command line.

'''Note''' FTP (insecure) is disabled on viper.hull.ac.uk and SFTP (secure FTP) is the recommended method.

===Large to a huge amount of data (100GB +)===
To avoid excessive network congestion and the long period to transfer such data please raise a [https://hull.service-now.com/ support ticket] for further details. This may typically involve connecting a USB data drive directly to VIPER to copy the data quickly into your user area.

===OneDrive to Viper===

With the University using MS OneDrive, it is possible to copy files to Viper. For multiple files, zip your files up on OneDrive first (considerably easier), as shown here:

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | From OneDrive to Viper
|-
|
* Starting with the zipped file MyData.zip on OneDrive, copied previously and compressed from local storage.
* Click on the '''share icon''' or options for the file here (three horizontal dots) and select '''share'''.

[[File:onedrive-01.png]]

* Then

[[File:onedrive-02.png]]

* Click on '''Copy'''

[[File:onedrive-03.png]]

* On your Viper console application, type '''wget''' and paste in the copied link and follow with '''-O yourfilename''' (otherwise Linux will name the file to the OneDrive link - ''argh!'')

[[File:onedrive-04.png]]

* And here is the received file

[[File:onedrive-05.png]]

'''Remember''' to un-share the file on OneDrive and use unzip (if required) on Linux to restore the file structure.

|}

==Universal==
The following can only be performed on Windows after installing MobaXterm or PuTTY.

The example user is user1; you will need to enter your username here:

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | * To Viper from local computer
|-
|scp myfile user1@viperlogin.hpc.hull.ac.uk:myfile
|-
|rsync -av my-directory viper.hull.ac.uk:/data/users/YourID
|}

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | * From Viper to your local computer
|-
| scp user1@viperlogin.hpc.hull.ac.uk:bio/YourUCInetID/hpc-file
|-
|rsync -av viper.hull.ac.uk:/data/users/YourID/mydatadirectory
|}

==Windows==
The following example uses MobaXterm, for transferring large files it is recommended to use FileZilla. For instructions on how to use FileZilla view the Linux section. follow the instruction in the Getting Connected section for installing MobaXterm.

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | * Uploading Files
|-
|* To upload files to Viper, navigate to the directory you wish to upload files to use the MobaXterm file browser.
* Choose the file/files to upload and click open or alternatively drag and drop the file/files you wish to upload over the Viper file browser.
* The upload may take some time depending on upload speed and file size.
|}

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | * Downloading Files
|-
|* To download files from Viper, navigate to the directory using the MobaXterm file browser.
[[File:mobaxterm_file_browser.jpg]]
* Select the file/files you wish to download and click the 'download selected files button'.
[[File:mobaxterm_download_icon.jpg]]
* Choose the directory where you want the files to download and click ok.
* The download will start, depending on download speed and file size this may take some time.
|}

==MacOS==
The following example uses Cyberduck, you can download Cyberduck from https://cyberduck.io macOS also supports FileZilla for an example of how to use FileZilla please view the Linux section.

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | Connecting
|-
|
* Click Open Connection.
* From the drop-down menu select '''SFTP''' (SSH File Transfer Protocol)
* Enter server viperlogin.hpc.hull.ac.uk, Port: 22, Username: University UID, Password: university password
* Click Connect.

|}

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | Uploading Files
|-
|
* Click Upload
* Choose the file/folder you want to upload and click Choose.
* The transfer window will appear.
* Please note due to file size the upload may take some time.
|}

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | Downloading Files
|-
|
* From the file browser right-click on the file/folder to download and select download from the menu.
* The transfer window will appear.
* Please note due to file size the download may take some time.
|}

==Linux==

The following example uses FileZilla, you can download FileZilla from https://filezilla-project.org/ (''Windows/Linux and MacOS'').
* '''Note''': the FileZilla installer sometimes comes with additional trial software packages. We would advise against installing these

{| class="wikitable mw-collapsible"
| style="width:100%; background: #aaaaaa;" | Using FileZilla
|-
|
*Enter the Host: viperlogin.hpc.hull.ac.uk, username, password, and port: 22, using an SFTP connection.
[[File:Filezilla_linux_login.jpg]]

* A pop-up will appear informing that the server host key is unknown. Click Always trust this host and click ok. In the remote site box, your Viper directory will appear.
* To transfer files click and drag the file/folder from either your pc to Viper (left to right) or from Viper to your pc (right to left). The transfer will start depending on transfer speed and file size this may take some time.

[[File:Filezilla_transferring_files.jpg]]

|}

==Navigation==

* [[Main_Page|Home]]
* [[Applications|Application support]]
* [[General|General]] *
* [[Programming|Programming support]]

Applications/Java

2024-09-20T20:35:58Z

Chris.collins: /* Application Details */

__TOC__

==Application Details==

* Description: Java is a general-purpose computer programming language that is concurrent, class-based, object-oriented,[14] and specifically designed to have as few implementation dependencies as possible. Only OpenJDK versions are available.
* Version: 6.5.14, 7.5.18 and 8.0.61
* Modules: openjdk/1.8.0_222-b10/gcc-8.2.0, openjdk/11.0.2/gcc-8.2.0, openjdk/openjdk-17.0.0_35
* Legacy Modules: java/jdk1.7.0_80 and java/jdk1.8.0_102 redirect to openjdk/1.8.0_222-b10/gcc-8.2.0
* Licence: Free, but owned by Oracle. Only OpenJDK versions are available and free and open source.

==Usage Examples==
{|
|style="width:5%; border-width: 0" | [[File:Icon_exclam3.png]]
|style="width:95%; border-width: 0" | Because Java is byte compiled and then run by a runtime program it requires the module to be loaded as well in a SLURM script. This is not the case with a standalone compiled program like from C/C++/Fortran source.
|-
|}

===Interactive Session===

'''Note''': this example is done on a compute node, usually access would be achieved with the scheduler

<pre style="background-color: black; color: white; border: 2px solid black; font-family: monospace, sans-serif;">

[username@gpu01 ~]$ module add java/jdk/1.8.0_102
[username@gpu01 ~]$ java myJAVA.javac

</pre>

===Batch Job===

This is running the job through the scheduler '''SLURM''' by using the command ''sbatch myJOB.job''

<pre style="font-family: monospace, sans-serif;">
#!/bin/bash

#SBATCH -J openJava-single-node
#SBATCH -N 1
#SBATCH --ntasks-per-node 1
#SBATCH -o %N.%j.%a.out
#SBATCH -e %N.%j.%a.err
#SBATCH -p compute
#SBATCH --exclusive

echo $SLURM_JOB_NODELIST

module purge
module add java/jdk1.8.0_102

export I_MPI_DEBUG=5
export I_MPI_FABRICS=shm:tmi
export I_MPI_FALLBACK=no

java /home/user/CODE_SAMPLES/demoJAVA

</pre>

==Further Information==

* [[Programming/Java|Java Programming Support (Viper)]]
* [https://www.java.com/en/ https://www.java.com/en/]

{{Modulepagenav}}

Applications/Java

2024-09-20T15:58:20Z

Chris.collins: /* Application Details */

__TOC__

==Application Details==

* Description: Java is a general-purpose computer programming language that is concurrent, class-based, object-oriented,[14] and specifically designed to have as few implementation dependencies as possible. Only OpenJDK versions are available.
* Version: 6.5.14, 7.5.18 and 8.0.61
* Modules: openjdk/1.8.0_222-b10/gcc-8.2.0, openjdk/11.0.2/gcc-8.2.0, openjdk/openjdk-17.0.0_35
* Legacy Modules: java/jdk1.7.0_80 and java/jdk1.8.0_102 redirect to openjdk/1.8.0_222-b10/gcc-8.2.0
* Licence: Free, but owned by Oracle

==Usage Examples==
{|
|style="width:5%; border-width: 0" | [[File:Icon_exclam3.png]]
|style="width:95%; border-width: 0" | Because Java is byte compiled and then run by a runtime program it requires the module to be loaded as well in a SLURM script. This is not the case with a standalone compiled program like from C/C++/Fortran source.
|-
|}

===Interactive Session===

'''Note''': this example is done on a compute node, usually access would be achieved with the scheduler

<pre style="background-color: black; color: white; border: 2px solid black; font-family: monospace, sans-serif;">

[username@gpu01 ~]$ module add java/jdk/1.8.0_102
[username@gpu01 ~]$ java myJAVA.javac

</pre>

===Batch Job===

This is running the job through the scheduler '''SLURM''' by using the command ''sbatch myJOB.job''

<pre style="font-family: monospace, sans-serif;">
#!/bin/bash

#SBATCH -J openJava-single-node
#SBATCH -N 1
#SBATCH --ntasks-per-node 1
#SBATCH -o %N.%j.%a.out
#SBATCH -e %N.%j.%a.err
#SBATCH -p compute
#SBATCH --exclusive

echo $SLURM_JOB_NODELIST

module purge
module add java/jdk1.8.0_102

export I_MPI_DEBUG=5
export I_MPI_FABRICS=shm:tmi
export I_MPI_FALLBACK=no

java /home/user/CODE_SAMPLES/demoJAVA

</pre>

==Further Information==

* [[Programming/Java|Java Programming Support (Viper)]]
* [https://www.java.com/en/ https://www.java.com/en/]

{{Modulepagenav}}

DAIM Guide

2024-06-10T10:46:44Z

Chris.collins: /* OnDemand Web Portal Access (pilot) */

__TOC__
''' Page currently under construction'''

== Introduction ==
Viper, the University of Hull’s High-Performance Computing facility, is used by research staff and students in many disciplines across the University and is a potentially significant tool for those with any sort of computational element to their research.
Viper is a ‘cluster’ of approximately 200 computes built into a cluster and features more than 6000 compute cores, high memory systems and GPU technology together with dedicated high-performance storage and fast interconnect between systems to meet the needs of the most computationally demanding tasks.
Some notes about
* Viper runs Linux, though to make use of Viper you do not need to be a Linux expert – being familiar with just a few commands is enough to get started.
* Viper runs a scheduler, a piece of software which manages access to the computers across the cluster, monitoring what is running and how resources such as memory, CPU processors and GPU cards are being used.
* Users run ‘jobs’ some of which are interactive (like Jupyter Notebook sessions) but many of which are submitted to run automatically without user interaction. Jobs usually have a batch script, a text file containing the recipe for the job, for example, what resource is needed and what the job should do.

== Prerequisites ==
* Viper account - to apply for a Viper account, please see [https://support.hull.ac.uk/tas/public/ssp/content/serviceflow?unid=c663e213f9364491bdcc801ccd809856 DAIM MSc Project Account Request]
* VPN - in order to access Viper from off-campus you will need a VPN account. Please see [https://support.hull.ac.uk/tas/public/ssp/content/serviceflow?unid=135c226e8cc7407ba312ac929d544d68 GlobalProtect Client VPN Service page on the Support Portal]

== Available Resource ==
'''Important''' While Viper has a range of GPU resources available for DAIM student use, including Nvidia A100 and Nvidia A40-based systems, such resource is in very high demand and as such at times there may be many people competing for GPU resource meaning long pend (queuing) times are likely.

Viper is a shared resource, so to help alleviate this for all users, could we suggest:

* Where possible, try and only run '''final''' GPU runs on Viper, and '''keep development of code to either other systems (e.g. own computer, DAIM lab computer or Viper CPU nodes)'''
* Don't request longer runtimes than is required or if you have resource allocated to you but are not going to work for a period of time ''' please cancel your allocation'''
* Please consider other users require GPU access, '''don't queue up multiple Jupyter sessions'''

{| class="wikitable" style="text-align:center;"
|- style="text-align:left;"
! Resource Type
! GPU RAM *
! System RAM
! Count
! Accessible as
! Notes
|-
| style="text-align:left;" | Nvidia A100 MIG 40GB
| 40GB
| 64GB
| 4
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | High demand
|-
| style="text-align:left;" | Nvidia A100 MIG 20GB
| 20GB
| 64GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | High demand
|-
| style="text-align:left;" | Nvidia A100 MIG 10GB
| 10GB
| 32GB
| 1
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | Medium demand

|-
| style="text-align:left;" | CPU only
| -
| 128GB / 1TB
| 10+
| style="text-align:left;" | [[DAIM_Guide#Pilot_Web_Portal_Access Ondemand | Web portal]]
| style="text-align:left;" | Lower demand but low performance for ML. Use for development and testing

|-
| style="text-align:left;" | NVidia A40
| 48GB
| 128GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A40_GPU_access | Web portal]]
| style="text-align:left;" | Very high demand, expect long wait times - these should only be used if the task exceeds GPU memory available on MIG instances
|}

<nowiki>*</nowiki> There are two specifications of DAIM lab PCs, the majority are equipped with Nvidia GeForce RTX 3070 GPU cards, with 8GB of GPU memory - all GPU instances on Viper have at least an equivalent amount of GPU memory available. There are also several lab PCs (dual monitor lab machines) with Nvidia RTX 3090 GPU cards with 24GB of GPU memory.

== OnDemand Web Portal Access (pilot) ==
We are currently testing a web route to provide Jupyter access via our pilot Viper OnDemand web portal - see http://hpc.mediawiki.hull.ac.uk/General/OOD

This service is in testing, please raise any issues via the University support portal, searching for '''Viper''' to find forms that will help you request support.

'''Please note:'''
* VPN access is still required to connect to the Viper OnDemand web portal.
* The DAIM Jupyter apps are in testing and there may be issues which have yet to be resolved.

If you connect to the web portal, you will see several DAIM prefixed apps, which can be used to connect:
==== Create Python Virtual Environment ====
* '''DAIM Create Environment''' - this will create a base Python virtual environment. You can change the name of the base environment by changing the text in the Environment Name box. You can then select this environment kernel to use when in DAIM session notebooks.

==== Web Portal Nvidia A40 GPU access ====
* '''DAIM Jupyter''' - this will launch a Jupyter lab session in your browser tab
* '''DAIM Jupyter Desktop''' - similarly, this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) access ====
* '''DAIM Jupyter Desktop MIG''' - this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.
* '''DAIM Jupyter MIG''' - this will launch a Jupyter lab session in your browser tab. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) automated run ====
* '''TEST DAIM nbconvert Job''' - This will submit a task to run Jupyter-nbconvert to automatically execute the specific notebook when the resource is available without you needing to interact with the notebook. Specify a notebook using the "select path" button and click launch. Once complete nbconvert will produce an output notebook with nbconvert in the name (e.g. the input mynotebook.ipynb will produce mynotebook.nbconvert.ipynb). After you launch the task, it will just run automatically when the resource is available without you needing to interact with it and will be listed as "Running" in your session list. Note that there is very little in terms of output until the notebook execution is complete. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal GPU and Job Status Page ====
To help work out what GPU resources are available for use, and how long a queued job may take, there is a monitor tool on the web portal accessible via '''Jobs > GPU Status and current jobs'''.

'''Warning:''' The first time this is run you will be presented with a warning about needing to restart your web server - doing this may terminate running Jupyter sessions, so do not proceed if you have active sessions.

Applications/Lumerical

2024-02-15T08:43:11Z

Chris.collins:

__TOC__
==Application Details==

* Description: Lumerical (Ansys)
* Version: 2023-R2
* Modules: lumerical/2023-R2/ompi
* Licence: Commercial

==Usage Examples==

3D/2D Maxwell's solver for nanophotonic devices and provides solutions for the following applications:

* CMOS Image Sensors
* Solar Cells
* OLEDs
* Integrated Optics
* Surface Metrology
* Surface Plasmons
* Metamaterials
* Photonic Crystals
* Liquid Crystals
* Graphene

'''Note''': connection to a FlexNet style license server will be required to run this software

In an interactive session, if you use the
<pre style="background-color: black; color: white; border: 2px solid black; font-family: monospace, sans-serif;">
singularitympiexec fdtd-engine-ompi-lcl ExampleLumericalTask.fsp </pre>
approach, this should try and make use of all the cores requested.

In a submission script, use:

<pre>
mpirun singularity exec ${LUMERICALAPP} fdtd-engine-ompi-lcl /home/USERNAME/ExampleLumericalTask.fsp
</pre>

E.g.

<pre>
#!/bin/bash
#SBATCH -J LumericalTest
#SBATCH -N 1
#SBATCH -n 40
#SBATCH -o %j.log
#SBATCH -e %j.err
#SBATCH -p ecompute
#SBATCH --time=10:00

module add lumerical/2023-R2/ompi

mpirun singularity exec ${LUMERICALAPP} fdtd-engine-ompi-lcl /home/USERNAME/ExampleLumericalTask.fsp
</pre>

(note this sample job runs for 10 minutes, please adjust as appropriate)

{{Licensepagenav}}

2023-10-19T08:48:34Z

Chris.collins:

Quickstart/Tutorial Videos

2023-10-19T08:47:28Z

Chris.collins: /* Where to find the tutorial videos */

==Where to find the tutorial videos==
We have created a number of tutorial videos. These short videos can be found either via MS Teams or MS Stream:

* In MS Teams, in the Teams tab on the top right is “Join or create team” button, from here either 1) In the “Join a team with a code” box enter '''k6gzwcc''' or use the “Search teams” and search for ISD and you should see a team called “ISD Viper HPC Users”. The tutorial videos can then be found under the "Tutorial Videos" tab on the main Teams screen.

* Alternatively the tutorial videos can be found on the Viper HPC Users Teams channel at [https://teams.microsoft.com/l/channel/19%3A297d547610284815b17aecb4398120a1%40thread.tacv2/tab%3A%3A4ae458ac-578d-4d73-abc2-a1c5d1f03f62?context=%7B%22channelId%22%3A%2219%3A297d547610284815b17aecb4398120a1%40thread.tacv2%22%7D&tenantId=490a8197-7b83-4f10-89b9-83189be3835e HPC Tutorial Videos on Stream ]

==What do the tutorial videos cover?==
These videos cover:
* An introduction to HPC
* Linux Fundamentals
* Linux Command Line
* Linux Tips
* Data on Viper
* An introduction to HPC
* Introduction to Slurm
* Using Modules
* Creating Batch Scripts
* Using R on Viper
* Using Python Virtual Environments
* Using Matlab on Viper

[[Main Page #Quickstart| Back]] / [[FurtherTopics/FurtherTopics| Move onto further topics]]

Quickstart/Tutorial Videos

2023-10-19T08:46:54Z

Chris.collins: /* Where to find the tutorial videos */

==Where to find the tutorial videos==
We have created a number of tutorial videos. These short videos can be found either via MS Teams or MS Stream:

* In MS Teams, in the Teams tab on the top right is “Join or create team” button, from here either 1) In the “Join a team with a code” box enter '''k6gzwcc''' or use the “Search teams” and search for ISD and you should see a team called “ISD Viper HPC Users”. The tutorial videos can then be found under the "Tutorial Videos" tab on the main Teams screen.

* Alternatively the tutorial videos can be found on the Viper HPC Users Teams channel at [https://teams.microsoft.com/l/channel/19%3A297d547610284815b17aecb4398120a1%40thread.tacv2/General?groupId=3d935f41-ef57-46a2-a961-cc5e30837c24&tenantId=490a8197-7b83-4f10-89b9-83189be3835e HPC Tutorial Videos on Stream ]

==What do the tutorial videos cover?==
These videos cover:
* An introduction to HPC
* Linux Fundamentals
* Linux Command Line
* Linux Tips
* Data on Viper
* An introduction to HPC
* Introduction to Slurm
* Using Modules
* Creating Batch Scripts
* Using R on Viper
* Using Python Virtual Environments
* Using Matlab on Viper

[[Main Page #Quickstart| Back]] / [[FurtherTopics/FurtherTopics| Move onto further topics]]

DAIM Guide

2023-10-18T09:56:01Z

Chris.collins: /* Available Resource */

__TOC__
''' Page currently under construction'''

== Introduction ==
Viper, the University of Hull’s High Performance Computing facility, is used by research staff and students in many disciplines across the University and is a potentially significant tool for those with any sort of computational element to their research.
Viper is a ‘cluster’ of approximately 200 computes built into a cluster, and features more than 6000 compute cores, high memory systems and GPU technology together with dedicated high performance storage and fast interconnect between systems to meet the needs of the most computationally demanding tasks.
Some notes about
* Viper runs Linux, though to make use of Viper you do not need to be a Linux expert – being familiar with just a few commands is enough to get started.
* Viper runs a scheduler, a piece of software which manages access to the computers across the cluster, monitoring what is running and how resources such as memory, cpu processors and GPU cards are being used.
* Users run ‘jobs’ some of which are interactive (like Jupyter Notebook sessions) but many of which are submitted to run automatically without user interaction. Jobs usually have a batch script, a text file containing recipe for the job, for example what resource is needed and what the job should do.

== Prerequisites ==
* Viper account - to apply for a Viper account, please see [https://support.hull.ac.uk/tas/public/ssp/content/serviceflow?unid=c663e213f9364491bdcc801ccd809856 DAIM MSc Project Account Request]
* VPN - in order to access Viper from off campus you will need a VPN account. Please see [https://support.hull.ac.uk/tas/public/ssp/content/serviceflow?unid=135c226e8cc7407ba312ac929d544d68 GlobalProtect Client VPN Service page on the Support Portal]

== Available Resource ==
'''Important''' While Viper has a range of GPU resources available for DAIM student use, including Nvidia A100 and Nvidia A40 based systems, such resource is in very high demand and as such at times there may be many people competing for GPU resource meaning long pend (queuing) times are likely.

Viper is a shared resource, so to help alleviate this for all users, could we suggest:

* Where possible, try and only run '''final''' GPU runs on Viper, and '''keep development of code to either other systems (e.g. own computer, DAIM lab computer or Viper CPU nodes)'''
* Don't request longer runtimes than is required or if you have resource allocated to you but are not going to work for a period of time '''please cancel your allocation'''
* Please consider other users require GPU access, '''don't queue up multiple Jupyter sessions'''

{| class="wikitable" style="text-align:center;"
|- style="text-align:left;"
! Resource Type
! GPU RAM *
! System RAM
! Count
! Accessible as
! Notes
|-
| style="text-align:left;" | Nvidia A100 MIG 20GB
| 20GB
| 64GB
| 9
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | High demand
|-
| style="text-align:left;" | Nvidia A100 MIG 10GB
| 10GB
| 32GB
| 3
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | Medium demand
|-
| style="text-align:left;" | CPU only
| -
| 128GB / 1TB
| 10+
| style="text-align:left;" | [[DAIM_Guide#Alternative_.28previous.29_method:_Jupyter_Notebook_via_SSH_tunnel_on_Viper | Viper SSH tunnel]]
| style="text-align:left;" | Lower demand but low performance for ML. Use for development and testing
|-
| style="text-align:left;" | CPU only
| -
| 128GB / 1TB
| 10+
| style="text-align:left;" | [[DAIM_Guide#Pilot_Web_Portal_Access Ondemand | Web portal]]
| style="text-align:left;" | Lower demand but low performance for ML. Use for development and testing
|-
| style="text-align:left;" | Nvidia A40
| 48GB
| 128GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Alternative_.28previous.29_method:_Jupyter_Notebook_via_SSH_tunnel_on_Viper | Viper SSH tunnel]]
| style="text-align:left;" | Very high demand, expect long wait times - these should only be used if task exceeds GPU memory available on MIG instances
|-
| style="text-align:left;" | NVidia A40
| 48GB
| 128GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A40_GPU_access | Web portal]]
| style="text-align:left;" | Very high demand, expect long wait times - these should only be used if task exceeds GPU memory available on MIG instances
|}

<nowiki>*</nowiki> There are two specifications of DAIM lab PCs, the majority are equipped with Nvidia GeForce RTX 3070 GPU cards, with 8GB of GPU memory - all GPU instances on Viper have at least an equivalent amount of GPU memory available. There are also a number of lab PCs (dual monitor lab machines) with Nvidia RTX 3090 GPU cards with 24GB of GPU memory.

== OnDemand Web Portal Access (pilot) ==
We are currently testing a web route to provide Jupyter access via our pilot Viper OnDemand web portal - see http://hpc.mediawiki.hull.ac.uk/General/OOD

This service is in testing, please raise any issues via the University [https://support.hull.ac.uk/tas/public/ssp/content/detail/service?unid=04cebf66128e4c8c8df023229c23b64a Support Portal]

'''Please note:'''
* VPN access is still required to connect to the Viper OnDemand web portal.
* The DAIM Jupyter apps are in testing and there may be issues which have yet to be resolved.

If you connect to the web portal, you will see a number of DAIM prefixed apps, which can be used to connect:
==== Create Python Virtual Environment ====
* '''DAIM Create Environment''' - this will create a base Python virtual environment. You can change the name of the base environment by changing the text in the Environment Name box. You can then select this environment kernel to use when in DAIM session notebooks.

==== Web Portal Nvidia A40 GPU access ====
* '''DAIM Jupyter''' - this will launch a Jupyter lab session in your browser tab
* '''DAIM Jupyter Desktop''' - similarly, this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) access ====
* '''TEST DAIM Jupyter Desktop MIG''' - this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.
* '''TEST DAIM Jupyter MIG''' - this will launch a Jupyter lab session in your browser tab. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) automated run ====
* '''TEST DAIM nbconvert Job''' - this will submit a task to run Jupyter-nbconvert to automatically execute the specific notebook when resource is available without you needing to interact with the notebook. Specify a notebook using the "select path" button and click launch. Once complete nbconvert will produce an output notebook with nbconvert in the name (e.g. the input mynotebook.ipynb will produce mynotebook.nbconvert.ipynb). After you launch the task, it will just run automatically when resource is available without you needing to interact with it and will be listed as "Running" in your session list. Note that there is very little in terms of output until the notebook execution is complete. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal GPU and Job Status Page ====
To help work out what GPU resources are available for use, and how long a queued job may take, there is a monitor tool on the web portal accessible via '''Jobs > GPU Status and current jobs'''.

'''Warning:''' the first time this is run you will be presented with a warning about needing to restart your webserver - doing this may terminate running Jupyter sessions, so do not proceed if you have active sessions.

== Alternative (previous) method: Jupyter Notebook via SSH tunnel on Viper ==
'''THIS APPROACH IS NO LONGER RECOMMENDED''':
A guide is available for step by step instructions for DAIM students to make use of Viper:
[http://hpc.mediawiki.hull.ac.uk/images/f/ff/Guide_to_using_Python_on_Viper_-_v1_3.pdf Guide to Jupyter Notebooks on Viper]

DAIM Guide

2023-10-18T08:16:04Z

Chris.collins: /* Alternative (previous) method: Jupyter Notebook via SSH tunnel on Viper */

__TOC__
''' Page currently under construction'''

== Introduction ==
Viper, the University of Hull’s High Performance Computing facility, is used by research staff and students in many disciplines across the University and is a potentially significant tool for those with any sort of computational element to their research.
Viper is a ‘cluster’ of approximately 200 computes built into a cluster, and features more than 6000 compute cores, high memory systems and GPU technology together with dedicated high performance storage and fast interconnect between systems to meet the needs of the most computationally demanding tasks.
Some notes about
* Viper runs Linux, though to make use of Viper you do not need to be a Linux expert – being familiar with just a few commands is enough to get started.
* Viper runs a scheduler, a piece of software which manages access to the computers across the cluster, monitoring what is running and how resources such as memory, cpu processors and GPU cards are being used.
* Users run ‘jobs’ some of which are interactive (like Jupyter Notebook sessions) but many of which are submitted to run automatically without user interaction. Jobs usually have a batch script, a text file containing recipe for the job, for example what resource is needed and what the job should do.

== Prerequisites ==
* Viper account - to apply for a Viper account, please see [https://support.hull.ac.uk/tas/public/ssp/content/serviceflow?unid=c663e213f9364491bdcc801ccd809856 DAIM MSc Project Account Request]
* VPN - in order to access Viper from off campus you will need a VPN account. Please see [https://support.hull.ac.uk/tas/public/ssp/content/serviceflow?unid=135c226e8cc7407ba312ac929d544d68 GlobalProtect Client VPN Service page on the Support Portal]

== Available Resource ==
'''Important''' While Viper has a range of GPU resources available for DAIM student use, including Nvidia A100 and Nvidia A40 based systems, such resource is in very high demand and as such at times there may be many people competing for GPU resource meaning long pend (queuing) times are likely.

Viper is a shared resource, so to help alleviate this for all users, could we suggest:

* Where possible, try and only run '''final''' GPU runs on Viper, and '''keep development of code to either other systems (e.g. own computer, DAIM lab computer or Viper CPU nodes)'''
* Don't request longer runtimes than is required or if you have resource allocated to you but are not going to work for a period of time '''please cancel your allocation'''
* Please consider other users require GPU access, '''don't queue up multiple Jupyter sessions'''

{| class="wikitable" style="text-align:center;"
|- style="text-align:left;"
! Resource Type
! GPU RAM *
! System RAM
! Count
! Accessible as
! Notes
|-
| style="text-align:left;" | Nvidia A40
| 48GB
| 128GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Alternative_.28previous.29_method:_Jupyter_Notebook_via_SSH_tunnel_on_Viper | Viper SSH tunnel]]
| style="text-align:left;" | Very high demand
|-
| style="text-align:left;" | NVidia A40
| 48GB
| 128GB
| 2
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A40_GPU_access | Web portal]]
| style="text-align:left;" | Very high demand
|-
| style="text-align:left;" | Nvidia A100 MIG 20GB
| 20GB
| 64GB
| 6
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | High demand
|-
| style="text-align:left;" | Nvidia A100 MIG 10GB
| 10GB
| 32GB
| 7
| style="text-align:left;" | [[DAIM_Guide#Web_Portal_Nvidia_A100_GPU_MIG_.28Multi-Instance_GPU.29_access | Web portal MIG version]]
| style="text-align:left;" | Medium demand
|-
| style="text-align:left;" | CPU only
| -
| 128GB / 1TB
| 10+
| style="text-align:left;" | [[DAIM_Guide#Alternative_.28previous.29_method:_Jupyter_Notebook_via_SSH_tunnel_on_Viper | Viper SSH tunnel]]
| style="text-align:left;" | Lower demand but low performance for ML. Use for development and testing
|-
| style="text-align:left;" | CPU only
| -
| 128GB / 1TB
| 10+
| style="text-align:left;" | [[DAIM_Guide#Pilot_Web_Portal_Access Ondemand | Web portal]]
| style="text-align:left;" | Lower demand but low performance for ML. Use for development and testing
|}

<nowiki>*</nowiki> There are two specifications of DAIM lab PCs, the majority are equipped with Nvidia GeForce RTX 3070 GPU cards, with 8GB of GPU memory - all GPU instances on Viper have at least an equivalent amount of GPU memory available. There are also a number of lab PCs (dual monitor lab machines) with Nvidia RTX 3090 GPU cards with 24GB of GPU memory.

== OnDemand Web Portal Access (pilot) ==
We are currently testing a web route to provide Jupyter access via our pilot Viper OnDemand web portal - see http://hpc.mediawiki.hull.ac.uk/General/OOD

This service is in testing, please raise any issues via the University [https://support.hull.ac.uk/tas/public/ssp/content/detail/service?unid=04cebf66128e4c8c8df023229c23b64a Support Portal]

'''Please note:'''
* VPN access is still required to connect to the Viper OnDemand web portal.
* The DAIM Jupyter apps are in testing and there may be issues which have yet to be resolved.

If you connect to the web portal, you will see a number of DAIM prefixed apps, which can be used to connect:
==== Create Python Virtual Environment ====
* '''DAIM Create Environment''' - this will create a base Python virtual environment. You can change the name of the base environment by changing the text in the Environment Name box. You can then select this environment kernel to use when in DAIM session notebooks.

==== Web Portal Nvidia A40 GPU access ====
* '''DAIM Jupyter''' - this will launch a Jupyter lab session in your browser tab
* '''DAIM Jupyter Desktop''' - similarly, this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) access ====
* '''TEST DAIM Jupyter Desktop MIG''' - this will launch a Jupyter session but with the web browser running on Viper OnDemand. This means you should be able to disconnect and reconnect with cells remaining active. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.
* '''TEST DAIM Jupyter MIG''' - this will launch a Jupyter lab session in your browser tab. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal Nvidia A100 GPU MIG (Multi-Instance GPU) automated run ====
* '''TEST DAIM nbconvert Job''' - this will submit a task to run Jupyter-nbconvert to automatically execute the specific notebook when resource is available without you needing to interact with the notebook. Specify a notebook using the "select path" button and click launch. Once complete nbconvert will produce an output notebook with nbconvert in the name (e.g. the input mynotebook.ipynb will produce mynotebook.nbconvert.ipynb). After you launch the task, it will just run automatically when resource is available without you needing to interact with it and will be listed as "Running" in your session list. Note that there is very little in terms of output until the notebook execution is complete. Options are between a 10GB and 20GB VRAM instance. Please only select what is required.

==== Web Portal GPU and Job Status Page ====
To help work out what GPU resources are available for use, and how long a queued job may take, there is a monitor tool on the web portal accessible via '''Jobs > GPU Status and current jobs'''.

'''Warning:''' the first time this is run you will be presented with a warning about needing to restart your webserver - doing this may terminate running Jupyter sessions, so do not proceed if you have active sessions.

== Alternative (previous) method: Jupyter Notebook via SSH tunnel on Viper ==
'''THIS APPROACH IS NO LONGER RECOMMENDED''':
A guide is available for step by step instructions for DAIM students to make use of Viper:
[http://hpc.mediawiki.hull.ac.uk/images/f/ff/Guide_to_using_Python_on_Viper_-_v1_3.pdf Guide to Jupyter Notebooks on Viper]