andre/blas-benchmarks
1
0
Fork 0
mirror of https://github.com/andre-wojtowicz/blas-benchmarks synced 2024-11-03 20:50:27 +01:00
Code Releases Activity
v1.0
blas-benchmarks/README.md

2282 lines
41 KiB
Markdown
Raw Permalink Normal View History

updated readme (first results)
2016-05-26 19:30:19 +02:00
# BLAS libraries benchmarks
Andrzej Wójtowicz
added OS info; added OC info for i7-4790K; fixed some legends (cuBLAS - work in progress)
2016-06-06 22:47:50 +02:00
Document generation date: 2016-06-06 22:40:18
updated readme (first results)
2016-05-26 19:30:19 +02:00
## Table of Contents
1. [Configuration](#configuration)
2. [Results per host](#results-per-host)
fixed typo
2016-06-06 15:10:29 +02:00
* [Intel Core i7-4790K + MSI GeForce GTX 980 Ti Lightning](#intel-core-i7-4790k--msi-geforce-gtx-980-ti-lightning)
fixed typo in ToC
2016-05-26 19:39:26 +02:00
* [Intel Core i5-4590 + NVIDIA GeForce GT 430](#intel-core-i5-4590--nvidia-geforce-gt-430)
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
* [Intel Core i5-4590 + NVIDIA GeForce GTX 750 Ti](#intel-core-i5-4590--nvidia-geforce-gtx-750-ti)
updated readme (first results)
2016-05-26 19:30:19 +02:00
* [Intel Core i5-3570](#intel-core-i5-3570)
* [Intel Core i3-2120](#intel-core-i3-2120)
* [Intel Core i3-3120M](#intel-core-i3-3120m)
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
* [Intel Core i5-3317U + NVIDIA GeForce GT 630M](#intel-core-i5-3317u--nvidia-geforce-gt-630m)
added Intel Pentium Dual-Core E5300
2016-06-03 15:40:33 +02:00
* [Intel Pentium Dual-Core E5300](#intel-pentium-dual-core-e5300)
updated readme (first results)
2016-05-26 19:30:19 +02:00
3. [Results per library](#results-per-library)
* [Netlib](#netlib)
* [Atlas (st)](#atlas-st)
* [OpenBLAS](#openblas)
* [Atlas (mt)](#atlas-mt)
* [GotoBLAS2](#gotoblas2)
* [MKL](#mkl)
* [BLIS](#blis)
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
* [cuBLAS](#cublas)
updated readme (first results)
2016-05-26 19:30:19 +02:00
***
## Configuration
added OS info; added OC info for i7-4790K; fixed some legends (cuBLAS - work in progress)
2016-06-06 22:47:50 +02:00
**OS**: Debian Jessie, kernel 4.4
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
**R software**: [Microsoft R Open](https://mran.microsoft.com/open/) (3.2.4)
updated readme (first results)
2016-05-26 19:30:19 +02:00
**Libraries**:
|CPU (single-threaded)|CPU (multi-threaded)|GPU|
|---|---|---|
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
|[Netlib](http://www.netlib.org/) (debian package, blas 1.2.20110419, lapack 3.5.0)|[OpenBLAS](http://www.openblas.net/) (debian package, 0.2.12)|[NVIDIA cuBLAS](https://developer.nvidia.com/cublas) (NVBLAS 6.5 + Intel MKL)|
|[ATLAS](http://math-atlas.sourceforge.net/) (debian package, 3.10.2)|[ATLAS](http://math-atlas.sourceforge.net/) (dev branch, 3.11.38)| |
| |[GotoBLAS2](https://prs.ism.ac.jp/~nakama/SurviveGotoBLAS2/) (Survive fork, 3.141)| |
| |[Intel MKL](https://mran.microsoft.com/download/) (part of RevoMath package, 3.2.4)| |
| |[BLIS](https://github.com/flame/blis) (dev branch, 0.2.0+/17.05.2016)| |
updated readme (first results)
2016-05-26 19:30:19 +02:00
**Hosts**:
|No.|CPU|GPU|
|---|---|---|
added OS info; added OC info for i7-4790K; fixed some legends (cuBLAS - work in progress)
2016-06-06 22:47:50 +02:00
|1.|[Intel Core i7-4790K](http://ark.intel.com/products/80807/Intel-Core-i7-4790K-Processor-8M-Cache-up-to-4_40-GHz) (OC 4.5 GHz)|[MSI GeForce GTX 980 Ti Lightning](https://us.msi.com/Graphics-card/GTX-980-Ti-LIGHTNING.html#hero-specification)|
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
|2.|[Intel Core i5-4590](http://ark.intel.com/products/80815/Intel-Core-i5-4590-Processor-6M-Cache-up-to-3_70-GHz)|[NVIDIA GeForce GT 430](http://www.geforce.com/hardware/desktop-gpus/geforce-gt-430/specifications)|
|3.|[Intel Core i5-4590](http://ark.intel.com/products/80815/Intel-Core-i5-4590-Processor-6M-Cache-up-to-3_70-GHz)|[NVIDIA GeForce GTX 750 Ti](http://www.geforce.com/hardware/desktop-gpus/geforce-gtx-750-ti/specifications)|
|4.|[Intel Core i5-3570](http://ark.intel.com/products/65702/Intel-Core-i5-3570-Processor-6M-Cache-up-to-3_80-GHz)| - |
|5.|[Intel Core i3-2120](http://ark.intel.com/products/53426/Intel-Core-i3-2120-Processor-3M-Cache-3_30-GHz)| - |
|6.|[Intel Core i3-3120M](http://ark.intel.com/products/71465/Intel-Core-i3-3120M-Processor-3M-Cache-2_50-GHz)| - |
|7.|[Intel Core i5-3317U](http://ark.intel.com/products/65707/Intel-Core-i5-3317U-Processor-3M-Cache-up-to-2_60-GHz)|[NVIDIA GeForce GT 630M](http://www.geforce.com/hardware/notebook-gpus/geforce-gt-630m/specifications)|
|8.|[Intel Pentium Dual-Core E5300](http://ark.intel.com/products/35300/Intel-Pentium-Processor-E5300-2M-Cache-2_60-GHz-800-MHz-FSB)| - |
updated readme (first results)
2016-05-26 19:30:19 +02:00
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
**Benchmarks**: [R-benchmark-25](http://r.research.att.com/benchmarks/R-benchmark-25.R), [Revolution](https://gist.github.com/andrie/24c9672f1ea39af89c66#file-rro-mkl-benchmark-r), [Gcbd](https://cran.r-project.org/web/packages/gcbd/vignettes/gcbd.pdf).
updated readme (first results)
2016-05-26 19:30:19 +02:00
# Results per host
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
## Intel Core i7-4790K + MSI GeForce GTX 980 Ti Lightning
updated readme (first results)
2016-05-26 19:30:19 +02:00
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h1_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h1_b3_t1.png)![](gen/img/img_ph_h1_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h1_b3_t2.png)![](gen/img/img_ph_h1_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h1_b3_t3.png)![](gen/img/img_ph_h1_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h1_b3_t4.png)![](gen/img/img_ph_h1_b3_t4b.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
## Intel Core i5-4590 + NVIDIA GeForce GT 430
updated readme (first results)
2016-05-26 19:30:19 +02:00
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h2_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h2_b3_t1.png)![](gen/img/img_ph_h2_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h2_b3_t2.png)![](gen/img/img_ph_h2_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h2_b3_t3.png)![](gen/img/img_ph_h2_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h2_b3_t4.png)![](gen/img/img_ph_h2_b3_t4b.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
## Intel Core i5-4590 + NVIDIA GeForce GTX 750 Ti
updated readme (first results)
2016-05-26 19:30:19 +02:00
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h3_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h3_b3_t1.png)![](gen/img/img_ph_h3_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h3_b3_t2.png)![](gen/img/img_ph_h3_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h3_b3_t3.png)![](gen/img/img_ph_h3_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h3_b3_t4.png)![](gen/img/img_ph_h3_b3_t4b.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
## Intel Core i5-3570
updated readme (first results)
2016-05-26 19:30:19 +02:00
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h4_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h4_b3_t1.png)![](gen/img/img_ph_h4_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h4_b3_t2.png)![](gen/img/img_ph_h4_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h4_b3_t3.png)![](gen/img/img_ph_h4_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h4_b3_t4.png)![](gen/img/img_ph_h4_b3_t4b.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
## Intel Core i3-2120
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h5_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h5_b3_t1.png)![](gen/img/img_ph_h5_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h5_b3_t2.png)![](gen/img/img_ph_h5_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h5_b3_t3.png)![](gen/img/img_ph_h5_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h5_b3_t4.png)![](gen/img/img_ph_h5_b3_t4b.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
## Intel Core i3-3120M
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
added new nvidia gpus and intel cpu; fixed gotoblas2 results; added package versions
2016-05-31 00:00:52 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h6_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h6_b3_t1.png)![](gen/img/img_ph_h6_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h6_b3_t2.png)![](gen/img/img_ph_h6_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h6_b3_t3.png)![](gen/img/img_ph_h6_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h6_b3_t4.png)![](gen/img/img_ph_h6_b3_t4b.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
## Intel Core i5-3317U + NVIDIA GeForce GT 630M
added Intel Pentium Dual-Core E5300
2016-06-03 15:40:33 +02:00
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
added Intel Pentium Dual-Core E5300
2016-06-03 15:40:33 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
added Intel Pentium Dual-Core E5300
2016-06-03 15:40:33 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
added Intel Pentium Dual-Core E5300
2016-06-03 15:40:33 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h7_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h7_b3_t1.png)![](gen/img/img_ph_h7_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h7_b3_t2.png)![](gen/img/img_ph_h7_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h7_b3_t3.png)![](gen/img/img_ph_h7_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h7_b3_t4.png)![](gen/img/img_ph_h7_b3_t4b.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
## Intel Pentium Dual-Core E5300
### R-benchmark-25
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b1_t1.png)
#### Linear regr. over a 2000x2000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b1_t2.png)
#### Eigenvalues of a 600x600 random matrix
BLIS hangs in this test
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_ph_h8_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h8_b3_t1.png)![](gen/img/img_ph_h8_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h8_b3_t2.png)![](gen/img/img_ph_h8_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h8_b3_t3.png)![](gen/img/img_ph_h8_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_ph_h8_b3_t4.png)![](gen/img/img_ph_h8_b3_t4b.png)
updated readme (first results)
2016-05-26 19:30:19 +02:00
# Results per library
## Netlib
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l1_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l1_b3_t1.png)![](gen/img/img_pl_l1_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l1_b3_t2.png)![](gen/img/img_pl_l1_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l1_b3_t3.png)![](gen/img/img_pl_l1_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l1_b3_t4.png)![](gen/img/img_pl_l1_b3_t4b.png)
## ATLAS (st)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l2_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l2_b3_t1.png)![](gen/img/img_pl_l2_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l2_b3_t2.png)![](gen/img/img_pl_l2_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l2_b3_t3.png)![](gen/img/img_pl_l2_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l2_b3_t4.png)![](gen/img/img_pl_l2_b3_t4b.png)
## OpenBLAS
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l3_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l3_b3_t1.png)![](gen/img/img_pl_l3_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l3_b3_t2.png)![](gen/img/img_pl_l3_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l3_b3_t3.png)![](gen/img/img_pl_l3_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l3_b3_t4.png)![](gen/img/img_pl_l3_b3_t4b.png)
## ATLAS (mt)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l4_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l4_b3_t1.png)![](gen/img/img_pl_l4_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l4_b3_t2.png)![](gen/img/img_pl_l4_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l4_b3_t3.png)![](gen/img/img_pl_l4_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l4_b3_t4.png)![](gen/img/img_pl_l4_b3_t4b.png)
## GotoBLAS2
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l5_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l5_b3_t1.png)![](gen/img/img_pl_l5_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l5_b3_t2.png)![](gen/img/img_pl_l5_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l5_b3_t3.png)![](gen/img/img_pl_l5_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l5_b3_t4.png)![](gen/img/img_pl_l5_b3_t4b.png)
## MKL
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l6_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l6_b3_t1.png)![](gen/img/img_pl_l6_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l6_b3_t2.png)![](gen/img/img_pl_l6_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l6_b3_t3.png)![](gen/img/img_pl_l6_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l6_b3_t4.png)![](gen/img/img_pl_l6_b3_t4b.png)
## BLIS
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
added Intel Pentium Dual-Core E5300
2016-06-03 15:40:33 +02:00
Intel Pentium Dual-Core E5300 hangs in this test
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l7_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l7_b3_t1.png)![](gen/img/img_pl_l7_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l7_b3_t2.png)![](gen/img/img_pl_l7_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l7_b3_t3.png)![](gen/img/img_pl_l7_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l7_b3_t4.png)![](gen/img/img_pl_l7_b3_t4b.png)
## cuBLAS
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
### R-benchmark-25
updated readme (first results)
2016-05-26 19:30:19 +02:00
#### 2800x2800 cross-product matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b1_t1.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Linear regr. over a 2000x2000 matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b1_t2.png)
added Intel Core i7-4790K and MSI GeForce GTX 980 Ti Lightning; fixed checkpoint libpath; fixed size of matrices in Urbanek (only in markdown); changed Urbanek to more common name R-benchmark-25 (only in markdown)
2016-06-06 15:07:17 +02:00
#### Eigenvalues of a 600x600 random matrix
updated readme (first results)
2016-05-26 19:30:19 +02:00
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b1_t3.png)
#### Determinant of a 2500x2500 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b1_t4.png)
#### Cholesky decomposition of a 3000x3000 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b1_t5.png)
#### Inverse of a 1600x1600 random matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b1_t6.png)
#### Escoufier's method on a 45x45 matrix
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b1_t7.png)
### Revolution benchmark
#### Matrix Multiply
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b2_t1.png)
#### Cholesky Factorization
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b2_t2.png)
#### Singular Value Deomposition
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b2_t3.png)
#### Principal Components Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b2_t4.png)
#### Linear Discriminant Analysis
Time in seconds - 10 runs - lower is better
![](gen/img/img_pl_l8_b2_t5.png)
### Gcbd benchmark
#### Matrix Multiply
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l8_b3_t1.png)![](gen/img/img_pl_l8_b3_t1b.png)
#### QR Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l8_b3_t2.png)![](gen/img/img_pl_l8_b3_t2b.png)
#### Singular Value Deomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l8_b3_t3.png)![](gen/img/img_pl_l8_b3_t3b.png)
#### Triangular Decomposition
Time in seconds regarding matrix size - right panel on log scale - from 50 to 5 runs - lower is better
![](gen/img/img_pl_l8_b3_t4.png)![](gen/img/img_pl_l8_b3_t4b.png)
initial commit
2016-04-28 13:38:54 +02:00
Copy Permalink