All the data presented in this paper was collected on the IBM-SP2 at UCLA's Office of Academic Computing, granted to UCLA by IBM Corporation under their Shared University Research Program.
Figure 2: Comparison of Predicted Execution Time of Collective I/O
Figure 3: Predicted Execution Time for Constant and Variable Size Cache
Figure 4: Number of Disk Read Ops for Constant and Variable Size Cache
Figure 5: Predicted Execution Time as a Function of IOnodes and Cnodes
Figure 6: Performance of Cache Policies for Matrix Multiplication
