Skip to content

Hawking

Hawking is the GPU supercalculator located in the ColdLAB.

Hawking's IP is listed on the Mattermost channel #Hawking, but you can also ask for it to the admins if you don't have access to the channel.

Creating an account

To create an account on Hawking, you need to ask the admins for it. Once you have an account, you can connect to Hawking with the following command: 

ssh user@ip_address

Connecting to Hawking

You can connect to hawking via ssh to run quickly any demanding piece of code. It's possible to do that directly in VScode with the SSH Extension.

Using Hawking with environments (required)

This machine provides a shared Miniforge installation located in: /opt/miniforge3

  • The base environment is shared and read-only
  • Users must never install packages in base
  • Each user has their own environments and package cache in $HOME/.conda

Rules (important)

Danger

DO

  • Create one Conda environment per project
  • Use mamba instead of conda when possible
  • Install packages inside environments only

DO NOT

  • Install packages in base
  • Use pip install in base
  • Mix pip and conda unless necessary

User guide

1. What is already configured

For every user:

  • conda and mamba are available in PATH
  • base does not auto-activate
  • User-local directories:

~/.conda/envs
~/.conda/pkgs
- Channels restricted to: conda-forge

Check anytime with:

conda info

2. How to create a new environment

Example: Python 3.11 environment 

mamba create -n myenv python=3.11
Activate it: 
conda activate myenv
Verify the environment is active: 
which python
Ensure it points to your user directory ~/.conda/envs/myenv/bin/python, not /opt/miniforge3.

3. Installing packages

Preferred (conda-forge) 

mamba install numpy scipy matplotlib
If a package is not available via Conda, use pip inside the activated environment: 
pip install some-package

Danger

Only inside an activated environment.

4. Managing environments

List your environments: 

conda env list
Remove an environment: 
conda remove -n myenv --all
Update an environment: 
mamba update -n myenv --all
Export environment (history only): 
conda env export --from-history > environment.yml
Recreate later: 
conda env create -f environment.yml

5. Typical “safe default” GPU environment

mamba create -n gpu-sci python=3.11
conda activate gpu-sci

mamba install \
  cuda-cudart cuda-cublas cuda-cufft cuda-curand \
  cupy pycuda pyfftw \
  numpy scipy matplotlib

Storage policy

Danger

Don't store anything on Hawking.
Before you get excited in the comms, and answer me “but yet I did put the path to LEON”: if you don't mount LEON there, it will just copy locally to Hawking!
This is what systematically crashes LEON (I went to look at the logs).
Every time you reboot, make sure you remount your LEON user on hawking with the ssh command 

sshfs **nom**@patriot.lkb.upmc.fr:/partages/EQ15B/LEON-15B /home/**USER**/LEON

If you have several GB used for your user account, this is usually a sign that you've started writing directly to hawking, you can check this by running 

du -h --max-depth=1
in your /home/user folder.

For admins

Account creation

To create an account for a new user, run the following command: 

sudo ./create_user_conda.sh alice
This will create a new user alice with a home directory and set up the necessary permissions for the shared Miniforge installation. The user will be able to create their own Conda environments and install packages without affecting other users.

Delete an account: 

sudo ./delete_user_conda.sh alice --yes

Machine documentation

Date: 2026-02-10T14:57:43+01:00

OS / Kernel

 Static hostname: hawking
       Icon name: computer-desktop
         Chassis: desktop
  Operating System: Ubuntu 24.04.3 LTS
          Kernel: Linux 6.8.0-86-generic
    Architecture: x86-64
 Hardware Vendor: Micro-Star International Co., Ltd.
  Hardware Model: MS-7D67
Firmware Version: 1.A2
   Firmware Date: Fri 2023-06-30
    Firmware Age: 2y 7month 1w 6d
Linux hawking 6.8.0-86-generic #87-Ubuntu SMP Distributor ID:   Ubuntu
Description:    Ubuntu 24.04.3 LTS
Release:    24.04
Codename:   noble

Hardware / CPU

Architecture:                         x86_64
CPU op-mode(s):                       32-bit, 64-bit
Address sizes:                        48 bits physical, 48 bits virtual
Byte Order:                           Little Endian
CPU(s):                               32
On-line CPU(s) list:                  0-31
Vendor ID:                            AuthenticAMD
BIOS Vendor ID:                       Advanced Micro Devices, Inc.
Model name:                           AMD Ryzen 9 7950X 16-Core Processor
BIOS Model name:                      AMD Ryzen 9 7950X 16-Core Processor             Unknown CPU @ 4.5GHz
BIOS CPU family:                      107
CPU family:                           25
Model:                                97
Thread(s) per core:                   2
Core(s) per socket:                   16
Socket(s):                            1
Stepping:                             2
CPU(s) scaling MHz:                   76%
CPU max MHz:                          5881.0000
CPU min MHz:                          400.0000
BogoMIPS:                             9000.32
Virtualization:                       AMD-V
L1d cache:                            512 KiB (16 instances)
L1i cache:                            512 KiB (16 instances)
L2 cache:                             16 MiB (16 instances)
L3 cache:                             64 MiB (2 instances)
NUMA node(s):                         1
NUMA node0 CPU(s):                    0-31