The system that was chosen is very similar to those found at other academic and government facilities. which will make it less difficult to port HPC code to the new FSU platform and vice-versa.  The system promises at least 2 TeraFlops of throughput in the first year and will easily grow given its' modular design.  

In year one the system will consist of the following components.

 

Computers and Storage:

• 4 Head nodes, [Dell PowerEdge 6950 severs] 
  
  • Quad, Dual-Core 2220 2.8 GHz Opterons, 2 GB RAM per core
• 128 Compute nodes [Dell PowerEdge SC1435 servers]
  
  • Dual, Dual-Core 2220 2.8 GHz Opterons, 2 GB RAM per core
  • Total of 512 Cores
• 90 TB of Panasas Parallel OSD Storage
  
  • 78 TB usable in a highly redundant raid configuration

Networking:  

• CISCO 6500 chassis for IP fabric
• CISCO 7024 Infiniband 288-port Chassis for MPI communication

Software:

• MOAB+TORQUE as the Scheduler and Resource Manager
• Linux CentOS (Free RedHat Enterprise Linux)

The following FSU faculty and staff members generously volunteered their time to serve on the HPC bid review panel; Eric Chassignet (COAPS), Paul Eugenio (Physics), Larry Conrad (OTI), Don Pace (OTC), Jeff McDonald (SCS), Jim Wilgenbusch (SCS), and  Phyllis Sullivan (Purchasing).