Home nodes:
Home nodes can be used for light shell access (e.g. things that were traditionally done on the legacy MCS login nodes), such as interactive sessions and no CPU-heavy jobs. You can directly SSH to these from any internal trusted network (desktops, Argonne-Auth wifi).
- homes.cels.anl.gov (round robin DNS entry for the hosts listed below)
- homes-01.cels.anl.gov
- homes-02.cels.anl.gov
Compute nodes:
These are for more CPU-heavy compute jobs. You can directly SSH to these from any internal trusted network (desktops, Argonne-Auth wifi). This list will grow as legacy compute nodes are rebuilt as GCE compute nodes.
- compute-01.cels.anl.gov (nomachine, see below)
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
Mem: 192GB
OS: Ubuntu 22.04
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
- compute-02.cels.anl.gov
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
Mem: 192GB
OS: Ubuntu 22.04
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
- compute-03.cels.anl.gov
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
Mem: 192GB
OS: Ubuntu 22.04
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
- compute-04.cels.anl.gov
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
Mem: 192GB
OS: Ubuntu 20.04
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
- compute-05.cels.anl.gov
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
Mem: 192GB
OS: Ubuntu 22.04
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
- compute-06.cels.anl.gov
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
Mem: 192GB
OS: Ubuntu 20.04
- CPU: 2X Intel Xeon Gold 6130 CPU @ 2.10GHz (16core / 32 threads each)
- compute-07.cels.anl.gov
- CPU: 1X AMD EPYC 7453 28-Core
Mem: 256GB
OS: Ubuntu 22.04
- CPU: 1X AMD EPYC 7453 28-Core
- compute-08.cels.anl.gov
- CPU: 2X Intel Ice Lake Xeon Gold 5317 @ 3.0GHz (12 core / 24 threads each)
GPU: Nvidia A30 24GB
Mem: 256GB
OS: Ubuntu 22.04
- CPU: 2X Intel Ice Lake Xeon Gold 5317 @ 3.0GHz (12 core / 24 threads each)
- compute-10.cels.anl.gov (compute-240-10) (big memory machine) (former compute001)
- CPU: 4X Intel(R) Xeon(R) CPU E7-4820 v2 @ 2.00GHz (8core / 16 threads each)
Mem: 1.5TB
OS: Ubuntu 20.04
- CPU: 4X Intel(R) Xeon(R) CPU E7-4820 v2 @ 2.00GHz (8core / 16 threads each)
- compute-11.cels.anl.gov
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
Mem: 64GB
OS: Ubuntu 20.04
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
- compute-12.cels.anl.gov
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
Mem: 64GB
OS: Ubuntu 20.04
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
- compute-13.cels.anl.gov
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
Mem: 64GB
OS: Ubuntu 20.04
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
- compute-14.cels.anl.gov
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
Mem: 64GB
OS: Ubuntu 20.04
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
- compute-15.cels.anl.gov
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
Mem: 64GB
OS: Ubuntu 22.04
- CPU: 2X Intel Xeon CPU E5-2650 0 @ 2.00GHz (8core / 16 threads each)
Other Architectures
- compute-macos-01.cels.anl.gov
- CPU: 1X Intel Core i7-8700B 3.20GHz
Mem: 64GB
OS: macOS 12.6.4
- CPU: 1X Intel Core i7-8700B 3.20GHz
- compute-macos-02.cels.anl.gov
- CPU: Apple M1 8-core CPU with 4 performance cores and 4 efficiency cores
8-core GPU
16-core Neural Engine
Mem: 16GB
OS: macOS 12.6.4
- CPU: Apple M1 8-core CPU with 4 performance cores and 4 efficiency cores
- compute-macos-03.cels.anl.gov
- CPU: Apple M1 Ultra 20-core with 16 performance cores and 4 efficiency cores
64-core GPU
32-core Neural Engine
Mem: 128GB
OS: macOS 13.3.1
- CPU: Apple M1 Ultra 20-core with 16 performance cores and 4 efficiency cores
Login nodes:
GCE login nodes function differently than the legacy MCS login nodes. While the login nodes are still your way in to the GCE environment while offsite, shell access on the login nodes is now prohibited. Instead, the login nodes will be used as an SSH proxy tunnel to your preferred GCE endpoint.
- logins.cels.anl.gov (round robin DNS entry for the login nodes listed below)
- login-01.cels.anl.gov
- login-02.cels.anl.gov
GUI nodes:
GCE also has a compute node available for interactive GUI logins via NoMachine. Download the client at nomachine.com and configure it to connect via SSH. If you are connecting from onsite or a trusted network, you can connect directly to nomachine.cels.anl.gov in the configuration.
Please note: By default, nomachine uses its own SSH instead of the system version, which ignores your .ssh/config. To fix this, you must edit your .nx/config/player.cfg file and change “SSH client mode” from library to native. See this for more information.
From offsite, you must either use the VPN or this SSH connection will have to be proxied through the logins. See SSH documentation.
- Quick cheat sheet:
- Mac/Linux users
- If you used the two line basic configuration from above, use the hostname “nomachine.cels.anl.gov” for the Host in the NoMachine setup.
- If you used the Advanced SSH config from above, add the following lines to your ~/.ssh/config immediately after the “homes-gce” section in the docs linked above.
Host nomachine nomachine.cels.anl.gov User <anlusername> Hostname nomachine.cels.anl.gov ProxyJump login-gce ForwardX11Trusted yes
Then, in nomachine, configure a connection via SSH to the host “nomachine” (presuming you used the example above).
- Important: Make sure in your connection, you go into the Configuration tab and specify Use key-based authentication with a SSH agent, then choose Modify and check Forward Authentication. If you do not do this, you may not be able to connect from outside Argonne networks.
- Windows users
- The Windows nxclient will not obey your .ssh config, so you must set up a separate SSH tunnel. To do this, put the following entries in your .ssh\config file:
Host nomachine nomachine.cels.anl.gov User <anlusername> Hostname nomachine.cels.anl.gov ProxyJump login-gce ForwardX11Trusted yes LocalForward 4100 localhost:4000
- Next, configure your connection to nomachine with the following settings:
- Name: CELS Nomachine
- Host: localhost
- Port: 4100
- Protocol: NX
- In the configuration tab for Authentication, Use password authentication.
- From a commandline, “ssh nomachine” and wait for a successful connection. After you see the shell prompt, then you can connect your nomachine client.
- The Windows nxclient will not obey your .ssh config, so you must set up a separate SSH tunnel. To do this, put the following entries in your .ssh\config file:
- Mac/Linux users
Server Fingerprints:
- login-01.cels.anl.gov
- MD5:f0:9a:f4:c8:86:be:69:8a:da:88:3d:1e:e0:f3:a3:be
- SHA256:LG4pRgxvfcAs5CjIWJFreDZHGX+G5rM8NFb4S5SF/aE
- login-02.cels.anl.gov
- MD5:f0:9a:f4:c8:86:be:69:8a:da:88:3d:1e:e0:f3:a3:be
- SHA256:LG4pRgxvfcAs5CjIWJFreDZHGX+G5rM8NFb4S5SF/aE
- homes-01.cels.anl.gov
- MD5:c5:70:a3:60:d4:35:a3:26:9c:eb:76:dc:1c:4f:46:7c
- SHA256:vjTFj4hLE4/iuwRABttTwLDWg867rhAXxpdqxPVTLko
- homes-02.cels.anl.gov
- MD5:c5:70:a3:60:d4:35:a3:26:9c:eb:76:dc:1c:4f:46:7c
- SHA256:vjTFj4hLE4/iuwRABttTwLDWg867rhAXxpdqxPVTLko
- compute-01.cels.anl.gov
- MD5:78:f9:10:9c:5f:97:65:cf:31:d0:2b:75:d8:40:c5:a7
- SHA256:Nskx6I5Ih34tImE7SFEBVZzC846bQ656tIYuICHZceM
- compute-02.cels.anl.gov
- MD5:3a:66:2c:f5:50:8c:17:d8:93:3e:95:97:26:42:5f:e4
- SHA256:iyWcecDeuF5fkP1YJgIMPcJOgFe3y5/ntdJIzvO6Iuc
- compute-03.cels.anl.gov
- MD5:68:2e:11:70:ce:32:48:19:09:b1:90:fb:9b:91:b8:f0
- SHA256:0ZGgsaHiBh6DSdet2MG7m+hoErFB9mAjxhnsQhdrCEQ
- compute-04.cels.anl.gov
- MD5:1b:ab:81:dc:fa:ad:91:b0:6a:66:01:cb:a6:29:0b:4f
- SHA256:syKWVcvLAsYg2KacvigfQHxvNT5FmnJwnKJFwPXrceU
- compute-05.cels.anl.gov
- MD5:4c:d0:95:72:32:c2:7d:77:9b:34:7f:1d:18:c0:c6:00
- SHA256:dAPuw6nA3X8+alWIimAK/ngFu6ZVk42WMBwO4Atuv5U
- compute-06.cels.anl.gov
- MD5:a6:ad:d9:ef:5d:dd:28:e6:79:42:75:2d:81:46:a8:0a
- SHA256:j6YCvR1doMi0VE93drkjy46NPg8CmyGfFGc8OKEtlPA
- compute-07.cels.anl.gov
- MD5:53:b3:ba:ae:3e:e6:79:b9:63:9b:83:1d:4d:ed:69:cd
- SHA256:xoGLJz/DxSaTz4+iVBqsbwBZgvU5NIxMhEaPEXEMmaI
- compute-10.cels.anl.gov
- MD5:4f:c3:a0:90:c4:ac:ad:8c:66:ea:0d:15:1a:fb:61:3c
- SHA256:exHjZFheSylNqGLIxWzKhkrApp/nOurey8c7UbaS6EA
- compute-11.cels.anl.gov
- MD5:60:e3:20:a1:34:d8:2e:c0:14:3b:b8:5c:1a:44:10:a3
- SHA256:jCa718F5od/wTbTWPhboPyzPD5bjxhlZAHS06toqDtE
- compute-12.cels.anl.gov
- MD5:a0:33:e3:34:cb:d0:1d:ee:9c:13:81:3a:62:0c:72:76
- SHA256:Ity96hpP/tQSI40pB+X61eHdsUtCOzur7n2KplbAS8k
- compute-13.cels.anl.gov
- MD5:0b:f2:4b:33:02:2e:c7:92:f8:1a:d6:f1:7f:43:a3:e3
- SHA256:rjk2SLoCo9k32gbSUN6i8jCy6iD4ln0vpQQr+IkMM0c
- compute-14.cels.anl.gov
- MD5:45:a1:e1:11:19:73:e8:de:8a:27:6b:b4:f0:96:29:a3
- SHA256:FGtiaZ8fJot7Oi7CnqcmYEztsjTttBqjPhadscQi4iQ
- compute-15.cels.anl.gov
- MD5:52:54:a7:ed:64:85:18:61:a3:51:53:60:7d:be:9f:c5
- SHA256:gOSEriaHa3gFkuwgbeHL469mMRTrBYRLITM5rMrOzMc