Purdue University

Retired Compute Resources

Purdue IT maintains many different resources for computation. Here is a brief introduction to current computational resources. More information and detailed documentation are available for each resource listed below.

BrownGPU

Brown-GPU was a new type of addition to Purdue's Community Clusters, designed specifically for applications which are able to take advantage of GPU accelerators. While applications must be specially-crafted to use GPUs, a GPU-enabled application can often run many times faster than the same application could on general-purpose CPUs. Due to the increased cost of GPU-equipped nodes, Brown-GPU was offered with some new purchase options to allow for shared access at a lower price point than the full cost of a node.

Brown

Brown is a Community Cluster optimized for communities running traditional, tightly-coupled science and engineering applications. Brown was built through a partnership with Dell and Intel in October 2017. Brown consists of Dell compute nodes with two 12-core Intel Xeon Gold "Sky Lake" processors (24 cores per node) and 96 GB of memory. All nodes have 100 Gbps EDR Infiniband interconnect and a 5-year warranty.

Workbench

The Data Workbench is an interactive compute environment for non-batch big data analysis and simulation, and is a part of Purdue's Community Cluster Program. The Data Workbench consists of HP compute nodes with two 8-core Intel Xeon processors (16 cores per node), and 128 GB of memory. All nodes are interconnected with 10 Gigabit Ethernet. The Data Workbench entered production on October 1, 2017.

HalsteadGPU

Halstead-GPU was a new type of addition to Purdue's Community Clusters, designed specifically for applications which are able to take advantage of GPU accelerators. While applications must be specially-crafted to use GPUs, a GPU-enabled application can often run many times faster than the same application could on general-purpose CPUs. Due to the increased cost of GPU-equipped nodes, Halstead-GPU was offered with some new purchase options to allow for shared access at a lower price point than the full cost of a node.

Halstead

Halstead is optimized for Purdue's communities running traditional, tightly-coupled science and engineering applications. Halstead was built through a partnership with HP and Intel in November 2016. Halstead consists of HP compute nodes with two 10-core Intel Xeon-E5 processors (20 cores per node) and 128 GB of memory. All nodes have 100 Gbps EDR Infiniband interconnect and a 5-year warranty.

Snyder

Snyder is a Purdue Community Cluster which is continually expanded and refreshed, optimized for data intensive applications requiring large amounts of shared memory per node, such as life sciences. Snyder was originally built through a partnership with HP and Intel in April 2015, though it has been most recently expanded with nodes from Dell. Snyder consists of a variety of compute node configurations as shown in the table below. All nodes have 40 Gbps Ethernet connections and a 5-year warranty. Snyder is expanded annually, with each year's purchase of nodes to remain in production for 5 years from their initial purchase.

Rice

Rice is optimized for Purdue's communities running traditional, tightly-coupled science and engineering applications. Rice was built through a partnership with HP and Intel in April 2015. Rice consists of HP compute nodes with two 10-core Intel Xeon-E5 processors (20 cores per node) and 64 GB of memory. All nodes have 56 Gbps FDR Infiniband interconnect and a 5-year warranty.

Conte

Conte was built through a partnership with HP and Intel in June 2013, and was the largest of Purdue's flagship community clusters. Conte consisted of HP compute nodes with two 8-core Intel Xeon-E5 processors (16 cores per node) and 64 GB of memory. Each node was also equipped with two 60-core Xeon Phi coprocessors. All nodes had 40 Gbps FDR10 Infiniband connections and a 5-year warranty. Conte was decommissioned on August 1, 2018.

Carter

Carter was launched through an ITaP partnership with Intel in November 2011 and was a member of Purdue's Community Cluster Program. Carter consisted of HP compute nodes with two 8-core Intel Xeon-E5 processors (16 cores per node) and between 32 GB and 256 GB of memory. A few NVIDIA GPU-accelerated nodes were also available. All nodes had 56 Gbps FDR Infiniband connections. Carter was decommissioned on April 30, 2017.

Hansen

Hansen was a compute cluster operated by ITaP and a member of Purdue's Community Cluster Program. Hansen went into production on September 15, 2011. Hansen consisted of Dell compute nodes with four 12-core AMD Opteron 6176 processors (48 cores per node) and either 96 GB, 192 GB, or 512 GB of memory. All nodes had 10 Gigabit Ethernet interconnects and a 5-year warranty. Hansen was decommissioned on October 1, 2016

Rossmann

Rossmann was a compute cluster operated by ITaP and was a member of Purdue's Community Cluster Program. Rossmann went into production on September 1, 2010. It consisted of HP (Hewlett Packard) ProLiant DL165 G7 nodes with 64-bit, dual 12-core AMD Opteron 6172 processors (24 cores per node) and 48 GB, 96 GB, or 192 GB of memory. All nodes had 10 Gigabit Ethernet interconnects and a 5-year warranty. Rossmann was decommissioned on November 2nd, 2015.

Coates

Coates was a compute cluster operated by ITaP and was a member of Purdue's Community Cluster Program. ITaP installed Coates on July 21, 2009, and at the time it was the largest entirely 10 Gigabit Ethernet (10GigE) academic cluster in the world. Coates consisted of 982 64-bit, 8-core Hewlett-Packard Proliant and 11 64-bit, 16-core Hewlett-Packard Proliant DL585 G5 systems with between 16 GB and 128 GB of memory. All nodes had 10 Gigabit Ethernet interconnects and a 5-year warranty. Coates was decommissioned on September 30, 2014.

Steele

Steele was a compute cluster operated by ITaP and the first system built under Purdue's Community Cluster Program. ITaP installed Steele in May 2008 in an unprecedented single-day installation. It replaced and expanded upon ITaP research resources retired at the same time, including the Hamlet, Lear, and Macbeth clusters. Steele consisted of 852 64-bit, 8-core Dell 1950 and 9 64-bit, 8-core Dell 2950 systems with various combinations of 16-32 GB RAM, 160 GB to 2 TB of disk, and 1 Gigabit Ethernet (1GigE) and InfiniBand local to each node.

WinHPC

WinHPC was a compute cluster operated by ITaP, and a member of Purdue's Community Cluster Program. WinHPC went into production on December 1, 2011. WinHPC consisted of HP compute nodes with two 12-core AMD Opteron 6172 processors (24 cores per node) and 48 GB of memory. All nodes had 10 Gigabit Ethernet interconnects and a 5-year warranty. WinHPC was decommissioned on October 1, 2016.

Miner

Miner was a compute cluster installed Miner at the Purdue Calumet campus on December 25, 2009 and operated by ITaP. It was the first major research cluster on the Calumet campus and represented a great step forward in Purdue Calumet's ongoing plan to foster more local, cutting-edge research. Miner consisted of 512 2-core Intel Xeon systems with either 4 or 6 GB RAM, 50 GB of disk, and 1 Gigabit Ethernet (1GigE) local to each node.

Gray

The Gray cluster was solely a development platform to be used alongside the Indiana Economic Development Corporation (IEDC) machine Black. Gray was a place to compile code (mostly serial Condor applications) that were to be run on Black. Black was housed with Indiana University's "Big Red" system in Bloomington, Indiana. However, Gray was located on Purdue's West Lafayette campus. Gray included a front-end server, several worker-node blades, and extra front-end hosts for campus and TeraGrid Condor users.

Black

The Black cluster was Purdue's portion of the Indiana Economic Development Corporation (IEDC) machine at Indiana University, the IU portion of which was known as 'Big Red'. Black consisted of 256 IBM JS21 Blades, each a Dual-Processor 2.5 GHz Dual-Core PowerPC 970 MP with 8 GB of RAM and PCI-X Myrinet 2000 interconnects. The large amount of shared memory in this system provided very fast communication between processor cores via shared memory and made the system ideal for large parallel jobs.

Gold

Gold was a small IBM Power5 system consisting of one front-end node and one compute node. The compute node was a Dual-Processor 1.5 GHz Dual-Core Power5 520 with 8 GB of RAM. This system was designed to be used only by users with legacy IBM architecture-optimized or AIX-specific code. Gold was intended to help facilitate the porting of any code designed for the IBM SP system at Purdue, which was retired in 2008.

Venice

Venice was a small cluster of Sun x4600 systems consisting of two front-end nodes and three compute nodes. The front-end nodes were both a Quad-Processor Dual-Core AMD Opteron 2216. The compute nodes were each an Eight-Processor Dual-Core AMD Opteron 8220 with 128 GB of RAM. The large amount of shared memory in this system made it ideal for large parallel jobs, using shared memory for fast communication between processor cores.

Prospero

The Prospero community cluster consisted of 19 Dell Quad-Processor 2.33 GHz Intel Xeon systems with 8 GB RAM and both Gigabit Ethernet and Infiniband interconnects. Each node had enough memory to run most jobs, and the high-speed Infiniband interconnect helped with many communication-bound parallel jobs.

Pete

The Pete cluster was composed of two parts, one owned by Earth, Atmospheric, and Planetary Science (EAPS) and the other by the Network for Computational Nanotechnology (NCN). Pete consisted of 166 HP Dual-Processor Dual-Core DL40 systems with either 8 or 16 GB RAM and Gigabit Ethernet. The large amount of memory in this system for its time made it well suited for larger-memory parallel jobs.

Moffett

Moffett was a SiCortex 5832 system. It consisted of 28 modules, each containing 27 six-processor SMP nodes for a total of 4536 processor cores. The SiCortex design was highly unusual; it paired relatively slow individual processor cores (633 MHz) with an extraordinarily fast custom interconnect fabric, and provided these in very large numbers. In addition, the SiCortex design used very little power and thereby generated very little heat.

Radon

Radon was a compute cluster operated by ITaP for general campus use. Radon's final iteration consisted of 45 HP Moonshot compute nodes with 32 GB of RAM that were connected by 10 Gigabit Ethernet (10GigE).

Hathi

Hathi was a shared Hadoop cluster operated by ITaP, and was a shared resource available to partners in Purdue's Community Cluster Program. Hathi went into production on September 8, 2014. Hathi consisted of 6 Dell compute nodes with two 8-core Intel E5-2650v2 CPUs, 32 GB of memory, and 48TB of local storage per node for a total cluster capacity of 288TB. All nodes had 40 Gigabit Ethernet interconnects and a 5-year warranty.

Peregrine1

Peregrine 1 was a state-of-the-art cluster for the Purdue Calumet campus operated by ITaP from the West Lafayette campus. Installed on June 26, 2012, Peregrine 1 was the second major research cluster to have been hosted on the Calumet campus. The cluster has been consequently relocated to the West Lafayette campus on August 19, 2015. Peregrine 1 consisted of HP compute nodes with two 8-core Intel Xeon-E5 processors (16 cores per node) and either 32 GB or 64 GB of memory. All nodes also featured 56 Gbps FDR Infiniband connections. Peregrine 1 was retired on October 12, 2016.