Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add benchmark for BLAS operations #772

Merged
merged 3 commits into from
May 26, 2021
Merged

Add benchmark for BLAS operations #772

merged 3 commits into from
May 26, 2021

Conversation

upsj
Copy link
Member

@upsj upsj commented May 22, 2021

This PR adds benchmarks for Ginkgo's BLAS-like operations (copy, axpy, scale, dot, norm, gemm) as well as a Dense::convert_to(Dense) implementation that doesn't modify the stride of the result vector if unnecessary (thus partially addressing #748). This benchmark can be used later to check whether #733 provides decent implementations for Dense operations.

The BLAS benchmarks are not the most sophisticated thing out there, since they don't flush the cache in any way, or check for correctness/precision (though the current design allows for that), but that should not matter too much as soon as we scale up.

A benchmark test case consists entirely of problem dimensions and strides.
We have input vectors x, y: nxr, a, b: 1xr or 1x1, A: nxk, B: kxm, C:nxm, for which we can specify individual strides stride_x/y/A/B/C with stride being a shorthand for stride_x/y.

The benchmark uses the amount of FLOPs and memory movement to compute FLOP/s and B/s numbers in addition to the runtime. Please double-check those (found in get_memory() and get_flops(), I just claimed that copy does one FLOP per entry), though.

Example output on my GPU:

[
    {
        "n": 1000000,
        "r": 10,
        "stride_x": 20,
        "stride_y": 15,
        "blas": {
            "copy": {
                "time": 0.019050621900000003,
                "flops": 524917246.92725118,
                "bandwidth": 8398675950.836019,
                "completed": true
            },
            "axpy": {
                "time": 0.0243559142,
                "flops": 821155791.3929588,
                "bandwidth": 9853869496.715506,
                "completed": true
            },
            "scal": {
                "time": 0.018665861,
                "flops": 535737408.5235072,
                "bandwidth": 8571798536.376115,
                "completed": true
            }
        }
    }
]

@upsj upsj added this to the Ginkgo 1.4.0 milestone May 22, 2021
@upsj upsj self-assigned this May 22, 2021
@upsj upsj added this to Awaiting Review in Ginkgo development May 22, 2021
@ginkgo-bot ginkgo-bot added mod:all This touches all Ginkgo modules. reg:benchmarking This is related to benchmarking. reg:build This is related to the build system. reg:testing This is related to testing. type:matrix-format This is related to the Matrix formats labels May 22, 2021
@upsj upsj added the 1:ST:ready-for-review This PR is ready for review label May 22, 2021
Copy link
Member

@yhmtsai yhmtsai left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

benchmark/blas/blas.cpp Show resolved Hide resolved
benchmark/blas/blas.cpp Show resolved Hide resolved
core/matrix/dense.cpp Outdated Show resolved Hide resolved
benchmark/blas/blas.cpp Show resolved Hide resolved
benchmark/blas/blas.cpp Outdated Show resolved Hide resolved
Co-authored-by: Yuhsiang Tsai <yhmtsai@gmail.com>
@codecov
Copy link

codecov bot commented May 25, 2021

Codecov Report

Merging #772 (aca6830) into develop (a1b57a0) will increase coverage by 0.01%.
The diff coverage is n/a.

Impacted file tree graph

@@             Coverage Diff             @@
##           develop     #772      +/-   ##
===========================================
+ Coverage    94.17%   94.18%   +0.01%     
===========================================
  Files          400      400              
  Lines        31049    31049              
===========================================
+ Hits         29241    29245       +4     
+ Misses        1808     1804       -4     
Impacted Files Coverage Δ
include/ginkgo/core/base/range.hpp 95.48% <ø> (ø)
omp/reorder/rcm_kernels.cpp 98.13% <0.00%> (+2.48%) ⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update a1b57a0...aca6830. Read the comment docs.

@sonarcloud
Copy link

sonarcloud bot commented May 25, 2021

Kudos, SonarCloud Quality Gate passed!

Bug A 0 Bugs
Vulnerability A 0 Vulnerabilities
Security Hotspot A 0 Security Hotspots
Code Smell A 13 Code Smells

0.0% 0.0% Coverage
0.0% 0.0% Duplication

@upsj upsj removed mod:all This touches all Ginkgo modules. reg:testing This is related to testing. labels May 25, 2021
Copy link
Member

@tcojean tcojean left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM.

I guess you don't want to advertise it which is why you didn't add it to neither run_all_benchmarks.sh or BENCHMARKING.md?

Ginkgo development automation moved this from Awaiting Review to Awaiting Merge May 26, 2021
@upsj upsj added 1:ST:ready-to-merge This PR is ready to merge. and removed 1:ST:ready-for-review This PR is ready for review 1:ST:run-full-test labels May 26, 2021
@upsj
Copy link
Member Author

upsj commented May 26, 2021

@tcojean Yes, I would like to finish some interface extensions (non-inplace or non-object-creating operations) and implementation changes (convert_to and move_to, mostly) first, so we can also represent conversions, transpose, permute etc. correctly.

@upsj upsj merged commit 95c7652 into develop May 26, 2021
@upsj upsj deleted the blas_benchmark branch May 26, 2021 14:50
tcojean added a commit that referenced this pull request Aug 20, 2021
Ginkgo release 1.4.0

The Ginkgo team is proud to announce the new Ginkgo minor release 1.4.0. This
release brings most of the Ginkgo functionality to the Intel DPC++ ecosystem
which enables Intel-GPU and CPU execution. The only Ginkgo features which have
not been ported yet are some preconditioners.

Ginkgo's mixed-precision support is greatly enhanced thanks to:
1. The new Accessor concept, which allows writing kernels featuring on-the-fly
memory compression, among other features. The accessor can be used as
header-only, see the [accessor BLAS benchmarks repository](https://github.com/ginkgo-project/accessor-BLAS/tree/develop) as a usage example.
2. All LinOps now transparently support mixed-precision execution. By default,
this is done through a temporary copy which may have a performance impact but
already allows mixed-precision research.

Native mixed-precision ELL kernels are implemented which do not see this cost.
The accessor is also leveraged in a new CB-GMRES solver which allows for
performance improvements by compressing the Krylov basis vectors. Many other
features have been added to Ginkgo, such as reordering support, a new IDR
solver, Incomplete Cholesky preconditioner, matrix assembly support (only CPU
for now), machine topology information, and more!

Supported systems and requirements:
+ For all platforms, cmake 3.13+
+ C++14 compliant compiler
+ Linux and MacOS
  + gcc: 5.3+, 6.3+, 7.3+, all versions after 8.1+
  + clang: 3.9+
  + Intel compiler: 2018+
  + Apple LLVM: 8.0+
  + CUDA module: CUDA 9.0+
  + HIP module: ROCm 3.5+
  + DPC++ module: Intel OneAPI 2021.3. Set the CXX compiler to `dpcpp`.
+ Windows
  + MinGW and Cygwin: gcc 5.3+, 6.3+, 7.3+, all versions after 8.1+
  + Microsoft Visual Studio: VS 2019
  + CUDA module: CUDA 9.0+, Microsoft Visual Studio
  + OpenMP module: MinGW or Cygwin.


Algorithm and important feature additions:
+ Add a new DPC++ Executor for SYCL execution and other base utilities
  [#648](#648), [#661](#661), [#757](#757), [#832](#832)
+ Port matrix formats, solvers and related kernels to DPC++. For some kernels,
  also make use of a shared kernel implementation for all executors (except
  Reference). [#710](#710), [#799](#799), [#779](#779), [#733](#733), [#844](#844), [#843](#843), [#789](#789), [#845](#845), [#849](#849), [#855](#855), [#856](#856)
+ Add accessors which allow multi-precision kernels, among other things.
  [#643](#643), [#708](#708)
+ Add support for mixed precision operations through apply in all LinOps. [#677](#677)
+ Add incomplete Cholesky factorizations and preconditioners as well as some
  improvements to ILU. [#672](#672), [#837](#837), [#846](#846)
+ Add an AMGX implementation and kernels on all devices but DPC++.
  [#528](#528), [#695](#695), [#860](#860)
+ Add a new mixed-precision capability solver, Compressed Basis GMRES
  (CB-GMRES). [#693](#693), [#763](#763)
+ Add the IDR(s) solver. [#620](#620)
+ Add a new fixed-size block CSR matrix format (for the Reference executor).
  [#671](#671), [#730](#730)
+ Add native mixed-precision support to the ELL format. [#717](#717), [#780](#780)
+ Add Reverse Cuthill-McKee reordering [#500](#500), [#649](#649)
+ Add matrix assembly support on CPUs. [#644](#644)
+ Extends ISAI from triangular to general and spd matrices. [#690](#690)

Other additions:
+ Add the possibility to apply real matrices to complex vectors.
  [#655](#655), [#658](#658)
+ Add functions to compute the absolute of a matrix format. [#636](#636)
+ Add symmetric permutation and improve existing permutations.
  [#684](#684), [#657](#657), [#663](#663)
+ Add a MachineTopology class with HWLOC support [#554](#554), [#697](#697)
+ Add an implicit residual norm criterion. [#702](#702), [#818](#818), [#850](#850)
+ Row-major accessor is generalized to more than 2 dimensions and a new
  "block column-major" accessor has been added. [#707](#707)
+ Add an heat equation example. [#698](#698), [#706](#706)
+ Add ccache support in CMake and CI. [#725](#725), [#739](#739)
+ Allow tuning and benchmarking variables non intrusively. [#692](#692)
+ Add triangular solver benchmark [#664](#664)
+ Add benchmarks for BLAS operations [#772](#772), [#829](#829)
+ Add support for different precisions and consistent index types in benchmarks.
  [#675](#675), [#828](#828)
+ Add a Github bot system to facilitate development and PR management.
  [#667](#667), [#674](#674), [#689](#689), [#853](#853)
+ Add Intel (DPC++) CI support and enable CI on HPC systems. [#736](#736), [#751](#751), [#781](#781)
+ Add ssh debugging for Github Actions CI. [#749](#749)
+ Add pipeline segmentation for better CI speed. [#737](#737)


Changes:
+ Add a Scalar Jacobi specialization and kernels. [#808](#808), [#834](#834), [#854](#854)
+ Add implicit residual log for solvers and benchmarks. [#714](#714)
+ Change handling of the conjugate in the dense dot product. [#755](#755)
+ Improved Dense stride handling. [#774](#774)
+ Multiple improvements to the OpenMP kernels performance, including COO,
an exclusive prefix sum, and more. [#703](#703), [#765](#765), [#740](#740)
+ Allow specialization of submatrix and other dense creation functions in solvers. [#718](#718)
+ Improved Identity constructor and treatment of rectangular matrices. [#646](#646)
+ Allow CUDA/HIP executors to select allocation mode. [#758](#758)
+ Check if executors share the same memory. [#670](#670)
+ Improve test install and smoke testing support. [#721](#721)
+ Update the JOSS paper citation and add publications in the documentation.
  [#629](#629), [#724](#724)
+ Improve the version output. [#806](#806)
+ Add some utilities for dim and span. [#821](#821)
+ Improved solver and preconditioner benchmarks. [#660](#660)
+ Improve benchmark timing and output. [#669](#669), [#791](#791), [#801](#801), [#812](#812)


Fixes:
+ Sorting fix for the Jacobi preconditioner. [#659](#659)
+ Also log the first residual norm in CGS [#735](#735)
+ Fix BiCG and HIP CSR to work with complex matrices. [#651](#651)
+ Fix Coo SpMV on strided vectors. [#807](#807)
+ Fix segfault of extract_diagonal, add short-and-fat test. [#769](#769)
+ Fix device_reset issue by moving counter/mutex to device. [#810](#810)
+ Fix `EnableLogging` superclass. [#841](#841)
+ Support ROCm 4.1.x and breaking HIP_PLATFORM changes. [#726](#726)
+ Decreased test size for a few device tests. [#742](#742)
+ Fix multiple issues with our CMake HIP and RPATH setup.
  [#712](#712), [#745](#745), [#709](#709)
+ Cleanup our CMake installation step. [#713](#713)
+ Various simplification and fixes to the Windows CMake setup. [#720](#720), [#785](#785)
+ Simplify third-party integration. [#786](#786)
+ Improve Ginkgo device arch flags management. [#696](#696)
+ Other fixes and improvements to the CMake setup.
  [#685](#685), [#792](#792), [#705](#705), [#836](#836)
+ Clarification of dense norm documentation [#784](#784)
+ Various development tools fixes and improvements [#738](#738), [#830](#830), [#840](#840)
+ Make multiple operators/constructors explicit. [#650](#650), [#761](#761)
+ Fix some issues, memory leaks and warnings found by MSVC.
  [#666](#666), [#731](#731)
+ Improved solver memory estimates and consistent iteration counts [#691](#691)
+ Various logger improvements and fixes [#728](#728), [#743](#743), [#754](#754)
+ Fix for ForwardIterator requirements in iterator_factory. [#665](#665)
+ Various benchmark fixes. [#647](#647), [#673](#673), [#722](#722)
+ Various CI fixes and improvements. [#642](#642), [#641](#641), [#795](#795), [#783](#783), [#793](#793), [#852](#852)


Related PR: #857
tcojean added a commit that referenced this pull request Aug 23, 2021
Release 1.4.0 to master

The Ginkgo team is proud to announce the new Ginkgo minor release 1.4.0. This
release brings most of the Ginkgo functionality to the Intel DPC++ ecosystem
which enables Intel-GPU and CPU execution. The only Ginkgo features which have
not been ported yet are some preconditioners.

Ginkgo's mixed-precision support is greatly enhanced thanks to:
1. The new Accessor concept, which allows writing kernels featuring on-the-fly
memory compression, among other features. The accessor can be used as
header-only, see the [accessor BLAS benchmarks repository](https://github.com/ginkgo-project/accessor-BLAS/tree/develop) as a usage example.
2. All LinOps now transparently support mixed-precision execution. By default,
this is done through a temporary copy which may have a performance impact but
already allows mixed-precision research.

Native mixed-precision ELL kernels are implemented which do not see this cost.
The accessor is also leveraged in a new CB-GMRES solver which allows for
performance improvements by compressing the Krylov basis vectors. Many other
features have been added to Ginkgo, such as reordering support, a new IDR
solver, Incomplete Cholesky preconditioner, matrix assembly support (only CPU
for now), machine topology information, and more!

Supported systems and requirements:
+ For all platforms, cmake 3.13+
+ C++14 compliant compiler
+ Linux and MacOS
  + gcc: 5.3+, 6.3+, 7.3+, all versions after 8.1+
  + clang: 3.9+
  + Intel compiler: 2018+
  + Apple LLVM: 8.0+
  + CUDA module: CUDA 9.0+
  + HIP module: ROCm 3.5+
  + DPC++ module: Intel OneAPI 2021.3. Set the CXX compiler to `dpcpp`.
+ Windows
  + MinGW and Cygwin: gcc 5.3+, 6.3+, 7.3+, all versions after 8.1+
  + Microsoft Visual Studio: VS 2019
  + CUDA module: CUDA 9.0+, Microsoft Visual Studio
  + OpenMP module: MinGW or Cygwin.


Algorithm and important feature additions:
+ Add a new DPC++ Executor for SYCL execution and other base utilities
  [#648](#648), [#661](#661), [#757](#757), [#832](#832)
+ Port matrix formats, solvers and related kernels to DPC++. For some kernels,
  also make use of a shared kernel implementation for all executors (except
  Reference). [#710](#710), [#799](#799), [#779](#779), [#733](#733), [#844](#844), [#843](#843), [#789](#789), [#845](#845), [#849](#849), [#855](#855), [#856](#856)
+ Add accessors which allow multi-precision kernels, among other things.
  [#643](#643), [#708](#708)
+ Add support for mixed precision operations through apply in all LinOps. [#677](#677)
+ Add incomplete Cholesky factorizations and preconditioners as well as some
  improvements to ILU. [#672](#672), [#837](#837), [#846](#846)
+ Add an AMGX implementation and kernels on all devices but DPC++.
  [#528](#528), [#695](#695), [#860](#860)
+ Add a new mixed-precision capability solver, Compressed Basis GMRES
  (CB-GMRES). [#693](#693), [#763](#763)
+ Add the IDR(s) solver. [#620](#620)
+ Add a new fixed-size block CSR matrix format (for the Reference executor).
  [#671](#671), [#730](#730)
+ Add native mixed-precision support to the ELL format. [#717](#717), [#780](#780)
+ Add Reverse Cuthill-McKee reordering [#500](#500), [#649](#649)
+ Add matrix assembly support on CPUs. [#644](#644)
+ Extends ISAI from triangular to general and spd matrices. [#690](#690)

Other additions:
+ Add the possibility to apply real matrices to complex vectors.
  [#655](#655), [#658](#658)
+ Add functions to compute the absolute of a matrix format. [#636](#636)
+ Add symmetric permutation and improve existing permutations.
  [#684](#684), [#657](#657), [#663](#663)
+ Add a MachineTopology class with HWLOC support [#554](#554), [#697](#697)
+ Add an implicit residual norm criterion. [#702](#702), [#818](#818), [#850](#850)
+ Row-major accessor is generalized to more than 2 dimensions and a new
  "block column-major" accessor has been added. [#707](#707)
+ Add an heat equation example. [#698](#698), [#706](#706)
+ Add ccache support in CMake and CI. [#725](#725), [#739](#739)
+ Allow tuning and benchmarking variables non intrusively. [#692](#692)
+ Add triangular solver benchmark [#664](#664)
+ Add benchmarks for BLAS operations [#772](#772), [#829](#829)
+ Add support for different precisions and consistent index types in benchmarks.
  [#675](#675), [#828](#828)
+ Add a Github bot system to facilitate development and PR management.
  [#667](#667), [#674](#674), [#689](#689), [#853](#853)
+ Add Intel (DPC++) CI support and enable CI on HPC systems. [#736](#736), [#751](#751), [#781](#781)
+ Add ssh debugging for Github Actions CI. [#749](#749)
+ Add pipeline segmentation for better CI speed. [#737](#737)


Changes:
+ Add a Scalar Jacobi specialization and kernels. [#808](#808), [#834](#834), [#854](#854)
+ Add implicit residual log for solvers and benchmarks. [#714](#714)
+ Change handling of the conjugate in the dense dot product. [#755](#755)
+ Improved Dense stride handling. [#774](#774)
+ Multiple improvements to the OpenMP kernels performance, including COO,
an exclusive prefix sum, and more. [#703](#703), [#765](#765), [#740](#740)
+ Allow specialization of submatrix and other dense creation functions in solvers. [#718](#718)
+ Improved Identity constructor and treatment of rectangular matrices. [#646](#646)
+ Allow CUDA/HIP executors to select allocation mode. [#758](#758)
+ Check if executors share the same memory. [#670](#670)
+ Improve test install and smoke testing support. [#721](#721)
+ Update the JOSS paper citation and add publications in the documentation.
  [#629](#629), [#724](#724)
+ Improve the version output. [#806](#806)
+ Add some utilities for dim and span. [#821](#821)
+ Improved solver and preconditioner benchmarks. [#660](#660)
+ Improve benchmark timing and output. [#669](#669), [#791](#791), [#801](#801), [#812](#812)


Fixes:
+ Sorting fix for the Jacobi preconditioner. [#659](#659)
+ Also log the first residual norm in CGS [#735](#735)
+ Fix BiCG and HIP CSR to work with complex matrices. [#651](#651)
+ Fix Coo SpMV on strided vectors. [#807](#807)
+ Fix segfault of extract_diagonal, add short-and-fat test. [#769](#769)
+ Fix device_reset issue by moving counter/mutex to device. [#810](#810)
+ Fix `EnableLogging` superclass. [#841](#841)
+ Support ROCm 4.1.x and breaking HIP_PLATFORM changes. [#726](#726)
+ Decreased test size for a few device tests. [#742](#742)
+ Fix multiple issues with our CMake HIP and RPATH setup.
  [#712](#712), [#745](#745), [#709](#709)
+ Cleanup our CMake installation step. [#713](#713)
+ Various simplification and fixes to the Windows CMake setup. [#720](#720), [#785](#785)
+ Simplify third-party integration. [#786](#786)
+ Improve Ginkgo device arch flags management. [#696](#696)
+ Other fixes and improvements to the CMake setup.
  [#685](#685), [#792](#792), [#705](#705), [#836](#836)
+ Clarification of dense norm documentation [#784](#784)
+ Various development tools fixes and improvements [#738](#738), [#830](#830), [#840](#840)
+ Make multiple operators/constructors explicit. [#650](#650), [#761](#761)
+ Fix some issues, memory leaks and warnings found by MSVC.
  [#666](#666), [#731](#731)
+ Improved solver memory estimates and consistent iteration counts [#691](#691)
+ Various logger improvements and fixes [#728](#728), [#743](#743), [#754](#754)
+ Fix for ForwardIterator requirements in iterator_factory. [#665](#665)
+ Various benchmark fixes. [#647](#647), [#673](#673), [#722](#722)
+ Various CI fixes and improvements. [#642](#642), [#641](#641), [#795](#795), [#783](#783), [#793](#793), [#852](#852)

Related PR: #866
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
1:ST:ready-to-merge This PR is ready to merge. reg:benchmarking This is related to benchmarking. reg:build This is related to the build system. type:matrix-format This is related to the Matrix formats
Projects
Ginkgo development
Awaiting Merge
Development

Successfully merging this pull request may close these issues.

None yet

4 participants