Pete Supercomputer
OSU's newest supercomputer "Pete" is available for all OSU researchers.
Pete’s computational infrastructure comprises 6,688 processor cores, 20,352 GB RAM, and more than 2 Petabytes of storage.
System summary:
-
164 compute nodes, each with dual Intel “Skylake” 6130 CPUs @ 2.1GHz and 96 GB RAM
-
36 compute nodes, each with dual Intel “Cascade Lake” 6226R CPUs @ 2.9GHz and 96 GB RAM
-
12 large memory nodes, each with dual Intel “Skylake” 6130 CPUs and 768 GB RAM
-
1 super memory node with dual Intel “Skylake” 6130 CPUs and 1.5 TB RAM
-
10 GPU nodes, each containing at least two NVIDIA Quadro RTX6000 GPUS with a total of 22 GPUs
-
100 Gb/s Intel Omni-path low-latency interconnect
-
2 PB high-speed parallel file system (for paid use)
-
360 TB high-speed parallel file system (/scratch)
-
Slurm Workload Manager
Codes Compiled on Pete
All fields of study at OSU, from physcal sciences and engineering, to mathematics and business, utilize Pete to discover innovative solulations for modern world questions. To aid
researchers in the this quest, HPCC staff, with the help of Pete users, have installed an extensive library of codes and software on Pete. A list of all the programs installed and
available to all users can be seen below.
Compilers:
gcc, intel, impi, openmpi, mpich, mvapich
Codes A-L:
abinit, afni, amber, anaconda, ansys, anvio, argtable, bcftools, bedops, bedtools, bismark, blasr, blast+, blobtools, boost, bowtie, bwa, canu, cd-hit, slustal-omega, cml, concoct, cplex, cuda, cufflinks, danps, dastool, diamond, discovardenovo, docker, eq-utils, emboss, energyplus, fastqc, fds, fftw, fraggenescan, freesurfer, gatk, gaussian, gdal, gdlib, go, grace, gromacs, gsl, gtdtk, gurobi, hdf5, hisat, hmmer, htslib, humann, hybpiper, idba, ior, iqtree, iseg, jags, jdk, jellyfish, jemalloc, jre, julia, kallisto, kat, king, kraken, libgd, libgdff, libxc
Codes M-Z:
madgraph, mathematica, matlab, maxbin, matebat, metaphlan, metasim, miniconda, minimap, mirdeep, mothur, mpfr, mumax, mummer, netcdf, nwchem, octopus, openfoam, pasa, pb-jelly, pbh5tools, perl, petsc, pplacer, prodigal, prophet, pullseg, python, qlime, qtlcartographer, quantum-espresso, R, raxml, rnammer, rsem, ruby, segemath, salmon, samtools, scalapack, siesta, singularity, snippy, solexaqa, spades, sratoolkit, stacks, star-ccm, star-rna, stringtie, structure, subread, tophat, transdecoder, trimgalore, trimmomatic, trinity, trinotate, vapor, velvet, vsearch, wget, yaml, zlib
OSU’s supercomputers and support team enable computational and data intensive research for hundreds of researchers throughout many of Oklahoma’s academic institutions and several national institutions.
The system was funded by MRI Award #1531128, Acquisition of Shared High Performance Compute Cluster for Multidisciplinary Computational and Data-Intensive Research.