Retired Clusters

Hudson

The Hudson cluster was  not decommissioned, but rather incorporated into the Greene cluster in April, 2023.
In 2020, AMD and technology partner Penguin Computing Inc donated AMD GPU nodes in order to support COVID-19 research. This included a donation to NYU, which became the Hudson AMD cluster. Hudson consisted of 20 compute nodes (servers), each equipped with an AMD EPYC Rome 7642 processor (having 48 processing cores), 512 gigabytes (GB) of host memory, 8 MI-50 32GB graphics processing units (GPUs) and 2 terabyte (TB) of local solid state disk (SSD) for data storage. As of April, 2023, these nodes have been incorporated into the Greene cluster in order to consolidate the expanding AMD computational ecosystem at NYU.

Peel

The Peel  Hadoop cluster was decommissioned on January 3rd, 2023. Peel was replaced by Dataproc on the Google Cloud Platform.

• Cloudera CDH 6.3.4 (Hadoop 3.0.0  with Yarn)

• 18 compute nodes

• CPU cores: 864 total (2x24-core per node)

• Total RAM: 6.9 TB (384 GB per node)

• HDFS total: 1.1 PB (64 TB per node)

Prince

The HPC Cluster Prince was decommissioned on February 1st, 2021 and replaced by the Greene cluster. 

• Prince was a production cluster for a majority of computing tasks, like general computations, numerical simulations, AI, and more.

• CPU cores: 10,084

• NVIDIA GPUs: 220

• Total RAM: 58 TB

• Disk space: 2.3 PB 

• Lustre and BeeGFS scratch file systems

• Infiniband interconnect

Dumbo

The Dumbo  Hadoop cluster was decommissioned on February 22nd, 2021. Dumbo was replaced by the Peel cluster.

• Cloudera CDH 5.15.0 (Hadoop 2.6.0  with Yarn)

• 44 compute nodes

• CPU cores: 704 total (2x8-core per node)

• Total RAM: 5.6 TB (128 GB per node)

• HDFS total: 1.4 PB

Brooklyn

The Brooklyn Research Cluster (BRC) was decommissioned on April 1st, 2021.

• OpenStack Software

• Used to create Bare metal machines, Virtual Machines (VMs), run containers, and Kubernetes clusters

• 25 servers

• CPU cores: 28 each server

• GPU NVIDIA: 4 each server

• RAM: 256 Gb each

• Disk space: 300 TB Ceph storage system

Mercer

The Mercer cluster was decommissioned on May 19th, 2017.