Izar#

Academic Cluster

This cluster is for Master students and courses. For pay-per-use accounts, you can check Kuma.

Useful info#

Connecting to the clusters#

To connect to the cluster you should:

ssh <username>@izar.hpc.epfl.ch

Here's the list of current fingerprints you should expect when connecting to this cluster:

ECDSA
    MD5:75:bf:66:87:94:39:94:93:9a:be:b5:06:2a:89:a4:67
    SHA256:CmnfF53AEyqgT0t0PfoRRxZ/3cLbanzMPZ4jcEmlItw
ED25519
    MD5:2a:9e:b4:bd:d5:2c:0c:7e:22:db:a9:b3:fb:ab:13:b6
    SHA256:q327bbvtHQTCLisv7au0d3rxw2JWO/+l3Gn2l0Z8AD4
RSA
    MD5:63:eb:8c:b5:d8:ef:2f:df:c1:9f:ab:85:c3:30:29:2c
    SHA256:JVMVBiqHiXVjcwzIpISxtxZHArjfYbBjoV6o3XUbkoA

QOS#

The standard QOS are:

gpu for jobs using up to 8 nodes, with a time limit of 3 days. This is the default;
long for jobs using up to 8 nodes, with a time limit of 7 days;
debug for testing you job on up to 2 GPUs (over two nodes if needed), with a time limit of 1 hour and with a high priority;
build for compiling codes, with up to 20 cores, 90 GB of RAM, zero GPUs, and a time limit of 4 hours.

Choose one with -q <qos> or --qos <qos>.

Partitions#

There are 3 partitions on Izar:

gpu, the default one, which can be used for most jobs;
gpu-xl, requests specifically access to the 2 XL nodes with 4 gpus each;
test, gives you access to the test node and can be used with the gpu QOS.

Choose one with -p <partition> or --partition <partition>.

Hardware characteristics#

This cluster has the following configuration:

Type	Count	Model	CPU	Memory	Storage	Naming	GPU #	GPU Model
Frontend	1	ProLiant XL190r Gen10 (U38)	Intel(R) Xeon(R) Gold 6230 CPU @ 2.10 GHz	196 GB	3.2 TB (NVMe)	izar1	NA	NA
Compute node	35	ProLiant XL190r Gen10 (U38)	Intel(R) Xeon(R) Gold 6230 CPU @ 2.10 GHz	196 GB	3.2 TB (NVMe)	i[01-35]	2	NVIDIA Corporation GV100GL Tesla V100 PCIe 32GB
Compute node	35	ProLiant XL190r Gen10 (U38)	Intel(R) Xeon(R) Gold 6230 CPU @ 2.10 GHz	384 GB	3.2 TB (NVMe)	i[36-70]	2	NVIDIA Corporation GV100GL Tesla V100 PCIe 32GB
Compute node	2	ProLiant XL270d Gen10 (U45)	Intel(R) Xeon(R) Gold 6230 CPU @ 2.10 GHz	768 GB	3.2 TB (NVMe)	ixl0[1,2]	4	NVIDIA Corporation GV100GL Tesla V100 SXM2 32GB
Admin server	1	ProLiant DL360 Gen10 Plus (U46)	Intel(R) Xeon(R) Gold 5317 CPU @ 3.00 GHz	64 GB	3.8 TB (NVMe)	iadmin1	NA	NA
Proxy server	1	Dell PowerEdge R630	Intel(R) Xeon(R) CPU E5-2620 v4 @ 2.10 GHz	64 GB	400 GB (SCSI)	scitas-internal-ismartproxy	NA	NA
Gateway server	1	HPE ProLiant DL380 Gen10 (U30)	Intel(R) Xeon(R) Gold 5215 CPU @ 2.50 GHz	128 GB	2.4 TB (SCSI)	igw	NA	NA