Neural Engineering Data Consortium logo

Neuronix: A Low-Cost High Performance Cluster

The Institute for Signal and Information Processing logo

The Neuronix high performance computing cluster allows us to conduct big data machine learning research. Some of our current projects being enabled by this cluster include:

  • AUTO_EEG: providing decision support to neurologists by automatically interpreting EEG signal events.
  • COHORT RETRIEVAL: identification of patients who are most similar to a target patient so that neurologists can better diagnose abnormalities in an EEG.
  • TUH_EEG: The world's largest publicly available EEG corpus representing 14 years of clinical EEGs collected at Temple University Hospital.

The Neuronix cluster was funded by the National Human Genome Research Institute of the NIH under award number U01HG008468ß.


Specifications:

  • File Server (1): 2x Intel Xeon (4 Cores) @ 3.0 GHz with 32 GB DDR4
  • Disk Space: 21 TB HDD, 0.5 TB SSD/node
  • Compute Nodes (4): 2x AMD Opteron (16 Cores) @ 2.4 with 128 GB DDR3
  • Operating System: CentOS
  • Software Control: Warwulf (provisioning), Torque (resource management), and Maui (scheduling).
  • Administration: Nagios (monitoring), Ganglia (data logging).
To learn more about the cluster, see our detailed description. Information about the hardware vendor can be found here.
A graph containing live CPU measurements from the NeuroNix Cluster. A photo of NeuroNix's chasis with its top removed. Network Diagram of Neuronix.

Footer
Up | Home | Courses | Projects | Proposals | Publications
Please direct questions or comments to joseph.picone@gmail.com