aboutsummaryrefslogtreecommitdiffhomepage
path: root/bench
Commit message (Collapse)AuthorAge
* syncGravatar Gael Guennebaud2010-07-07
|\
* | add support for complexGravatar Gael Guennebaud2010-07-07
| |
* | add intitial support for the vectorization of complex<float>Gravatar Gael Guennebaud2010-07-05
| |
| * check for !x86 platforms, otherwise the BTL benchmark doesn't compile on ↵Gravatar Konstantinos Margaritis2010-07-05
|/ | | | arm/powerpc
* Fix cache computation on old Intel CPUs which do notGravatar Gael Guennebaud2010-06-27
| | | | support the cpuid function 0x4
* add the manual Intel's way to query cache infoGravatar Gael Guennebaud2010-06-26
|
* add a utilility to debug cpuid, and makes sure we get 0 if we query an ↵Gravatar Gael Guennebaud2010-06-26
| | | | unsupported cpuid function
* email changeGravatar Gael Guennebaud2010-06-24
|
* add support for oskiGravatar Gael Guennebaud2010-06-24
|
* btl: add a trmm action and update eigen interfaceGravatar Gael Guennebaud2010-06-23
|
* add a spmv mini becnhmark for Eigen, GMM++, ublas, mtl4, and oskiGravatar Gael Guennebaud2010-06-22
|
* slightly optimize computeProductBlockingSizes by explicitely precomputing ↵Gravatar Gael Guennebaud2010-06-22
| | | | what is known at compile time
* fix in case we don't know how to query the L1/L2 cache sizesGravatar Gael Guennebaud2010-06-21
|
* simplify and optimize block sizes computation for matrix products. TheyGravatar Gael Guennebaud2010-06-21
| | | | | are now automatically computed from the L1 and L2 cache sizes which are themselves automatically determined at runtime.
* make bench_gemm print out the queried cache sizesGravatar Gael Guennebaud2010-06-21
|
* add the possibility to set the cache size at runtimeGravatar Gael Guennebaud2010-06-18
|
* add runtime API to control multithreadingGravatar Gael Guennebaud2010-06-10
|
* make BenchTimer compatible with 2.0 branchGravatar Gael Guennebaud2010-06-01
|
* remove USING_PART_OF_NAMESPACE_EIGEN, leaving it in Eigen2Support.Gravatar Benoit Jacob2010-04-22
| | | | improve porting-Eigen2-to-3 docs
* Fixed line endings.Gravatar Hauke Heibel2010-03-05
|
* add a small program to bench all combinations of small productsGravatar Gael Guennebaud2010-03-05
|
* clean a bit the bench_gemm filesGravatar Gael Guennebaud2010-03-05
|
* minor cleaningGravatar Gael Guennebaud2010-03-05
|
* merge with default branchGravatar Gael Guennebaud2010-03-04
|\
| * clean #defined tokens, and use clock_gettime for the real timeGravatar Gael Guennebaud2010-03-03
| |
| * BenchTimer: avoid warning about symbol redefinition on win32, and include ↵Gravatar Eamon Nerbonne2010-03-02
| | | | | | | | <Eigen/Core> (required to compile)
* | remove Qt's atomic dependency, I don't know what I was doing wrong...Gravatar Gael Guennebaud2010-03-01
| |
* | make Aron's idea work using Qt's atomic implementation for the synchronisationGravatar Gael Guennebaud2010-03-01
| |
* | BTL: allow to bench real timeGravatar Gael Guennebaud2010-02-26
| |
* | fix some BTL issuesGravatar Gael Guennebaud2010-02-26
| |
* | implement a smarter parallelization strategy for gemm avoiding multipleGravatar Gael Guennebaud2010-02-26
| | | | | | | | paking of the same data
* | update BTL (better timer, eigen2 => eigen3, etc)Gravatar Gael Guennebaud2010-02-23
| |
| * mergeGravatar Gael Guennebaud2010-02-22
| |\
| | * provide default values for CXX, remove duplicate defineGravatar Thomas Capricelli2010-02-22
| | |
| | * upsGravatar Hauke Heibel2010-02-22
| | |
* | | fix BTL's eigen interfaceGravatar Gael Guennebaud2010-02-22
| | | | | | | | | | | | | | | (transplanted from 437f40acc1cbd9ce2f2a2a3f413cae3a5b35f8fb )
* | | significant speedup in the matrix-matrix productsGravatar Gael Guennebaud2010-02-23
| | |
* | | oopsGravatar Gael Guennebaud2010-02-22
| | |
* | | Port BenchTimer fix.Gravatar Hauke Heibel2010-02-22
| | |
* | | mergeGravatar Gael Guennebaud2010-02-22
|\ \ \ | | |/ | |/|
| * | Added getRealTime() for windows.Gravatar Hauke Heibel2010-02-22
| | |
* | | add a small benchmark to quickly bench/compare SMP supportGravatar Gael Guennebaud2010-02-22
|/ /
* | extend the bench timer to allow benchmarking of parallel code,Gravatar Gael Guennebaud2010-02-22
| | | | | | | | improvements are welcome
| * fix BTL's eigen interfaceGravatar Gael Guennebaud2010-02-22
|/
* mergeGravatar Mark Borgerding2010-02-16
|\
* | added benchmark for unscaled and half-spectrum FFTsGravatar Mark Borgerding2010-01-21
| |
| * extend sparse product benchmark with ublasGravatar Gael Guennebaud2010-02-09
|/
* extend benchmark for sparse productsGravatar Gael Guennebaud2010-01-05
|
* Big renaming:Gravatar Benoit Jacob2010-01-04
| | | | | | start ---> head end ---> tail Much frustration with sed syntax. Need to learn perl some day.
* * Fix bug #79: ei_alignmentOffset was assuming that ptr is multiple ofGravatar Benoit Jacob2010-01-02
| | | | | | | sizeof(Scalar), and that assumption breaks with double on linux x86-32. * Rename ei_alignmentOffset to ei_first_aligned * Rewrite its documentation and part of its body * The variant taking a MatrixBase doesn't need a separate size argument.