Benchmarks of the Paper: Conflict-Directed Delay-Bounding for Analyzing GPU Programs with Atomics


1  Our Benchmark Set

Benchmarkcorrectness-checkingbuggybugfree
nbodyprogrammer assertionsherehere
nbodyfinal-state comparisonherehere
tspprogrammer assertionsherehere
tspfinal-state comparisonherehere
aMinprogrammer assertionherehere
aMinfinal-state comparisonherehere
aMinUpdateprogrammer assertionsherehere
aMinUpdatefinal-state comparisonherehere
btreeprogrammer assertionsherehere
btreefinal-state comparisonherehere

2  Other Downloads

2.1  GKLEE

The prototype tool, GKLEEatm, is coming but not available right now. We will integrate into GKLEE in the near future. But you can try GKLEE here.

2.2  The Original N-body and Traveling Salesmen Kernels

The original CUDA n-body and traveling salesman kernels are from Prof. Martin Burtscher.

You can find the origainl CUDA kernel here.

The Traveling Salesman kernel is here.


This document was translated from LATEX by HEVEA.