Adaptive Checkpointing With Bandwidth-Aware Optimization in HPC [Data set]

Data collected for the thesis “Adaptive Checkpointing With Bandwidth-Aware Optimization in HPC”. Instructions are provided in the TMIO GitHub repository: https://github.com/tuda-parallel/TMIO/blob/development/docs/checkpoint_reproducibility.md

The contents are structured as follows:

Data
├── VELOC_BANDWIDTH
│     └── ... // Contains traces for VELOC checkpoint interval by bandwidth      
├── WACOMM_FREQUENCY
│     └── ... // Contains traces for frequency as I/O phase substitute     
├── HACC_IO_RESULTS 
│     └── ... // Contains traces for fine-grained bandwidth limit approaches     
├── PREFETCH
│     └── ... // Contains traces for I/O frequency based prefetching
├── HACC_IO_GROWING_PHASES
│     └── ... // Contains traces for fine-grained bandwidth limit 
│             // approaches with growing bandwidth requirements 
├── HACC_IO_VARYING_PHASES
│     └── ... // Contains traces for fine-grained bandwidth limit 
│             // approaches with varying bandwidth requirements during iterations 
├── HETDIS_VELOC
│     └── ... // Contains traces for asynchronous checkpoint evaluation
└── HEATDIS_ITERATIONS
     └── ... // Contains traces for asynchronous slowdown factor examination