Difference: GPU_inspiral (1 vs. 4)

Revision 42011-08-04 - GergelyDebreczeni

Line: 1 to 1
 
META TOPICPARENT name="WebHome"

GPU_inspiral

Added:
>
>
 GPU_inspiral is a low-latency, high performance many-core implementation of the matched-filter gravitational wave search algorithm, developed by the RMKI Virgo Group. The base is an architecture and vendor independent OpenCL code. The sub algorithms implemented so far are the following:
Line: 11 to 12
 
  • chi2 calculation
  • clustering of events
Added:
>
>		The implementation makes use of the stream processing paradigm applied to the matched filter problem. Various atomic operations are mapped to one or some of the kernel classes like map, reduce, fold, etcc... which are deeply studied and examined already in many-core architecture programming. The complete pipeline runs on the GPU, no CPU - GPU communication happens, which gives an additional acceleration of the code.. The implementation is straightforward to extend to multi-GPU or GPU-cluster environment,

 

Performance

Added:
>
>
 Performence test was done on a 2048 sec long data chunk, 500 template of 64 sec long was filtered. The code run on an Nvidia Tesla C2050. To produce the SNR time series it took ~17 sec
Changed:
<
<		for gpu_inpiral, this has to be compare with the ~40 minutes necessary for lalapps_inspiral to produce the same result.
>
>		for gpu_inpiral, this has to be compare with the ~40 minutes necessary for lalapps_inspiral to produce the same result.
 
Changed:
<
<

This results a *speed-up factor of 2 orders of magnitude !

>
>

This results a speed-up factor of 2 orders of magnitude !


 

Documentation

Line: 28 to 46
 
Added:
>
>		 Complete description and documentation in english available soon !
 

Future plans

Instead of writing up a completely new analysis software we will incorporate the code developed as gpu_inspiral to be part of the pyCBC project.

Developers

Added:
>
>
  1. Gergely Debreczeni (Gergely.Debreczeni@rmki.kfki.hu)
 
  1. Bence Somhegyi
Deleted:
<
<
  1. Gergely Debreczeni
 \ No newline at end of file

Revision 32011-06-30 - GergelyDebreczeni

Line: 1 to 1
 
META TOPICPARENT name="WebHome"

GPU_inspiral

Line: 24 to 24
 
  • A poster about the project here.
  • A summary paper here.
  • Bence's Somhegyi MSc thesis (in hungarian) here.
Changed:
<
<
  • Bence's Somhegyi MSc thesis presentation (in hungarian) here.
>
>
 

Future plans

Instead of writing up a completely new analysis software we will incorporate the code developed as gpu_inspiral

Revision 22011-06-28 - GergelyDebreczeni

Line: 1 to 1
 
META TOPICPARENT name="WebHome"

GPU_inspiral

Line: 20 to 20
 

Documentation

Changed:
<
<		Ther is some material available on this project:
>
>		There is some material available on this project:
  • A poster about the project here.
  • A summary paper here.
  • Bence's Somhegyi MSc thesis (in hungarian) here.
  • Bence's Somhegyi MSc thesis presentation (in hungarian) here.
 

Future plans

Added:
>
>		Instead of writing up a completely new analysis software we will incorporate the code developed as gpu_inspiral to be part of the pyCBC project.
 

Developers

  1. Bence Somhegyi

Revision 12011-06-28 - GergelyDebreczeni

Line: 1 to 1
Added:
>
>
META TOPICPARENT name="WebHome"

GPU_inspiral

GPU_inspiral is a low-latency, high performance many-core implementation of the matched-filter gravitational wave search algorithm, developed by the RMKI Virgo Group. The base is an architecture and vendor independent OpenCL code. The sub algorithms implemented so far are the following:

  • template generation
  • PSD calculation
  • matched-filtering
  • chi2 calculation
  • clustering of events

Performance

Performence test was done on a 2048 sec long data chunk, 500 template of 64 sec long was filtered. The code run on an Nvidia Tesla C2050. To produce the SNR time series it took ~17 sec for gpu_inpiral, this has to be compare with the ~40 minutes necessary for lalapps_inspiral to produce the same result.

This results a *speed-up factor of 2 orders of magnitude !

Documentation

Ther is some material available on this project:

Future plans

Developers

  1. Bence Somhegyi
  2. Gergely Debreczeni
View topic | History: r4 < r3 < r2 < r1 | More topic actions...
 
This site is powered by the TWiki collaboration platform Powered by Perl This site is powered by the TWiki collaboration platformCopyright &© by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback