1001n Research Cluster¶

The center provides virtualized research computing services through a Hyper-Converged Infrastructure (HCI) cluster called 1001n (a thousand and one nights!), named after the famous Arabian folk story 1001 Nights. The cluster provides virtualized computing resources through a VMware vSAN environment built from 5 VMware-ready nodes.

The cluster nodes are grouped in two groups; Compute Only (Sinbad) nodes and Compute/ GPU nodes (Aladdin) nodes. The Sinbad series has a total of 4 nodes and the Aladdin series has a total of 1 node. Each Sinbad node has 512 Gb Memory, 36 Cores, 21 TB (6. 99 TB x 3 disks) NVMe disks and a 1.46 TB (write intensive) NVMe disks. The Aladdin node has 2 x Nvidia Tesla V100 (32 GB) GPU cards, 358 GB of memory, 36 Cores, 14 TB (6.99 TB x 2 disks) NVMe disks and a 1.75 TB (write Intensive) NVMe disk.

The 1001n cluster in total has 2 TB of memory, 97.81 TB of storage and 180 CPU Cores which can support over 700 vCPUs. The nodes are connected through redundant high-performance switches. Uplink bandwidth to the rest of the campus network is 10 GB/s and the cluster also has a dedicated link with Dalma, the HPC Cluster for direct access to work storage.

1001n Research Cluster¶

Social Sciences Nodes¶