IntelPython/dpbench

Issues

• Make it possible to build dpbench without having to install and build all included implementations

  #220 opened 8 months ago by diptorupd
  1

• Add experimental to --npbench and --polybench

  #298 opened 8 months ago by ZzEeKkAa
  0

• Run numba_mlir benchmarks only with --experimental-numba-mlir flag

  #297 opened 8 months ago by ZzEeKkAa
  0

• Build and run instructions for CUDA devices

  #338 opened a year ago by rickybalin
  1

• Add wheels upload to dppy/label/dev

  #295 opened 8 months ago by ZzEeKkAa
  1

• Analyze performance issues with rodinia/gaussian implementation

  #312 opened a year ago by ZzEeKkAa
  0

• Implement Rodinia Back Propagantion benchmark

  #311 opened 10 months ago by ZzEeKkAa
  1

• Add pytest infrastructure for dpbench

  #238 opened a year ago by adarshyoga
  0

• Reorganize benchmarks

  #240 opened a year ago by diptorupd
  5

• Adding JIT compilation time of benchmarks to dpbench report

  #336 opened a year ago by rickybalin
  0

• Crash in child process causes entire runner to terminate with exception

  #282 opened a year ago by Hardcode84
  1

• Add numba-mlir and numba-mlir replace_parfors npbench benchmarks

  #292 opened a year ago by Hardcode84
  0

• DBscan dpex_k implementation does not work under compute follows data.

  #140 opened 2 years ago by diptorupd
  0

• Validator should be a member of a benchmark and not a framework

  #156 opened a year ago by diptorupd
  5

• Sporadic CI failures

  #320 opened a year ago by AlexanderKalistratov
  2

• Rodinia Gaussian benchmark does not work for fp32 sycl

  #316 opened a year ago by ZzEeKkAa
  0

• WARNING:root: Precision unchanged for " + arg_name + " due to unsupported type.

  #315 opened a year ago by npolina4
  0

• Implement single precision data support for npbench.

  #314 opened a year ago by npolina4
  0

• Update pairwise distance sycl implementation

  #290 opened a year ago by adarshyoga
  0

• Add CHANGELOG.md

  #296 opened a year ago by ZzEeKkAa
  0

• Add way to exclude workload from run

  #286 opened 2 years ago by AlexanderKalistratov
  0

• Improve benchmark results data model

  #285 opened 2 years ago by ZzEeKkAa
  0

• Improvements to dpbench CI, reporting, and other infrastructure improvements

  #184 opened 2 years ago by diptorupd
  0

• Print a table with the platform/device the benchmark ran.

  #187 opened 2 years ago by ZzEeKkAa
  0

• Provide a way to run benchmarks with alternate framework configs

  #244 opened 2 years ago by diptorupd
  1

• SYCL workloads failing on windows

  #267 opened 2 years ago by Hardcode84
  4

• Only python and numpy implementations available as reference

  #271 opened 2 years ago by AlexanderKalistratov
  1

• Package structure and dependencies does not allow strait type linting

  #186 opened 2 years ago by ZzEeKkAa
  1

• dataclass_transform() got an unexpected keyword argument 'field_specifiers'

  #251 opened 2 years ago by Hardcode84
  3

• dpbench fails if no reference implementation for benchmark

  #272 opened 2 years ago by AlexanderKalistratov
  1

• Benchmarks hang when executed on GPU

  #166 opened 2 years ago by adarshyoga
  3

• Improvements to implementation summary output to make it more informative

  #145 opened 2 years ago by diptorupd
  2

• Possible improvements to framework

  #123 opened 2 years ago by diptorupd
  3

• Occsional hang when running benchmarks

  #119 opened 2 years ago by diptorupd
  4

• Gpairs numba_dpex_k fails validation with preset = "M"

  #126 opened 2 years ago by diptorupd
  1

• Make framework selection extensible

  #221 opened 2 years ago by Hardcode84
  0

• CI should flag expected failures

  #234 opened 2 years ago by ZzEeKkAa
  0

• Turn on the validation step in Github CI

  #120 opened 2 years ago by diptorupd
  3

• Disabling validation should not be reported as a validation failure

  #211 opened 2 years ago by diptorupd
  0

• Add Dpnp implementations to dpbench

  #124 opened 2 years ago by diptorupd
  1

• L2 distance benchmark fails validation

  #127 opened 2 years ago by diptorupd
  1

• kmeans dpex_n is based on numba prange

  #138 opened 2 years ago by diptorupd
  1

• pairwise_distance validation fail on pipeline with MKL error

  #155 opened 2 years ago by mingjie-intel
  3

• Report problem size in MB/GB rather than preset values such as "S", "M"

  #144 opened 2 years ago by diptorupd
  0

• Add an automation to generate the __init__.py file for each benchmark

  #118 opened 2 years ago by diptorupd
  1

• dpbench crashing with numba-dpex 0.19.0

  #169 opened 2 years ago by mingjie-intel
  2

• Resetting input args across repeated executions of workloads fails to reset all args

  #147 opened 2 years ago by adarshyoga
  0

• pairwise_distance dpnp impl fails validation

  #161 opened 2 years ago by DenisScherbakov
  1

• Properly handle workloads crashes and hungs

  #154 opened 2 years ago by Hardcode84
  0

• Disable pairwise_distance_numba_dpex_n.py

  #121 opened 2 years ago by mingjie-intel
  0