benchmark results repository

[musm]

Feel free to post results of running the benchmarks suite, including versioninfo() and commit number for reference

[ChrisRackauckas]

High-Powered Dual Xeon(R) CPU E5-2667 v4 @ 3.20GHz:
https://gist.github.com/ChrisRackauckas/fb2c187317af097c61f03e7d99b71bb2

[ChrisRackauckas]

AMD FX-8350:
https://gist.github.com/ChrisRackauckas/97906a7ead17da9b08377bccec552af0

[ChrisRackauckas]

Overclocked i7 4770K:
https://gist.github.com/ChrisRackauckas/68a8a77f4e9cca42a611e8e2e4a0c534

[musm]

As always thanks @ChrisRackauckas . I addressed the perf of the inverse trig functions and exp# functions, and hopefully a bit of the cbrt.

[KristofferC]

https://gist.github.com/KristofferC/99d9eec3b4c6f7522cf8334306063316

julia> versioninfo()
Julia Version 0.6.0-dev.736
Commit 32c1a31* (2016-09-21 21:04 UTC)
Platform Info:
  System: Linux (x86_64-linux-gnu)
  CPU: Intel(R) Core(TM) i5-2500K CPU @ 3.30GHz
  WORD_SIZE: 64
  BLAS: libopenblas (USE64BITINT DYNAMIC_ARCH NO_AFFINITY Sandybridge)
  LAPACK: libopenblas64_
  LIBM: libopenlibm
  LLVM: libLLVM-3.7.1 (ORCJIT, sandybridge)

commit: d2b90b6

[musm]

Thanks @KristofferC I just merged #43 which I hope will bring most things to within 2.5

[musm]

Nonetheless results don't look bad at all, I'm hoping JuliaLang/julia#18632 will also help perf, but my understanding is that simple immutables are already stack allocated, but perhaps not inlined into arrays ?

[ChrisRackauckas]

Yes, immutables are already stack allocated for simple immutables. Maybe that helps Complex?

Running a new one right now.

[ChrisRackauckas]

New Xeon Benchmark:
https://gist.github.com/ChrisRackauckas/84ecbc26a244f0319b6a85c7e3f539e1

[KristofferC]

That pr should do nothing for performance here.

[musm]

Thanks for the clarification @KristofferC , that's what I figured.

[pfitzseb]

i5-4570
https://gist.github.com/pfitzseb/6c0851e3fb41c8565d425b088e4aaa04