timTrack tracking of charged particles By J.A.Rodríguez
  TRAS G O  Project lab CAF
<ul><li>RPC  TRB  timTrack </li></ul>
timTrack Resultados.txt (output file) - 6  SAETA  (x,y,x',y',v,t) - 6  Errors   -15  Covariances running ... Datos.txt Det...
<ul><li>Why  C   language  ?  </li></ul><ul><ul><li>Very fast </li></ul></ul><ul><ul><li>Flexible </li></ul></ul><ul><ul><...
BLAS /LAPACK <ul><li>Is a software library for numerical linear algebra. </li></ul><ul><li>It provides routines for solvin...
Intel® IPP <ul><li>Integrated Performance Primitives (Intel® IPP) </li></ul><ul><li>Is a library of multicore-ready, highl...
timTrack  SAETAs solutions <ul><li>PREVIOUS VERSION </li></ul><ul><li>timTrack v1.0  (LAPACK) </li></ul><ul><li>timTrack v...
timTrack   variance-covariance matrix <ul><li>PREVIOUS VERSIONS </li></ul><ul><li>timTrack v1.0  (LAPACK) </li></ul><ul><l...
Example Implemented Z Y T1 T2 X
Times for 1.000.000 particles Old  Python and Matlab versions   (only 500.000 particles)     165m  47.137 s  timTrack v2.0...
Next Steps <ul><li>Analyze systematic computing errors </li></ul><ul><li>Check single-precision version </li></ul><ul><li>...
timTrack v2.1 Next step  ( still in progress… )  Parallelims with  Intel® MPI  libraries Shared parallelism with OpenMP fo...
Future !  <ul><li>timtrack v 3.0   </li></ul><ul><li>CUDA  parallel computing architecture in GPUs  </li></ul><ul><li>CUDA...
 
Upcoming SlideShare
Loading in...5
×

A Gomez T Tat Cesga

348

Published on

0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total Views
348
On Slideshare
0
From Embeds
0
Number of Embeds
0
Actions
Shares
0
Downloads
1
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide

A Gomez T Tat Cesga

  1. 1. timTrack tracking of charged particles By J.A.Rodríguez
  2. 2.   TRAS G O Project lab CAF
  3. 3. <ul><li>RPC TRB timTrack </li></ul>
  4. 4. timTrack Resultados.txt (output file) - 6 SAETA (x,y,x',y',v,t) - 6 Errors -15 Covariances running ... Datos.txt Detector.txt
  5. 5. <ul><li>Why C language ? </li></ul><ul><ul><li>Very fast </li></ul></ul><ul><ul><li>Flexible </li></ul></ul><ul><ul><li>Parallelism </li></ul></ul><ul><ul><li>A rich set of libraries </li></ul></ul><ul><li>Libraries was used to program timTrack (“algorithms ”) </li></ul><ul><li>LAPACK </li></ul><ul><li>Intel® IPP </li></ul>
  6. 6. BLAS /LAPACK <ul><li>Is a software library for numerical linear algebra. </li></ul><ul><li>It provides routines for solving systems of linear equations and linear least squares, eigenvalue problems, and singular value decomposition. </li></ul><ul><li>Specific versions for each CPU model provided by the vendors </li></ul>
  7. 7. Intel® IPP <ul><li>Integrated Performance Primitives (Intel® IPP) </li></ul><ul><li>Is a library of multicore-ready, highly optimized software functions for digital media and data-processing applications. </li></ul><ul><li>Intel IPP contains a rich set of matrix and vector operations for a wide variety of applications. </li></ul>
  8. 8. timTrack SAETAs solutions <ul><li>PREVIOUS VERSION </li></ul><ul><li>timTrack v1.0 (LAPACK) </li></ul><ul><li>timTrack v1.1 (IPP) </li></ul><ul><li>NEW algebra VERSION </li></ul><ul><li>timTrack v2.0 (LAPACK) </li></ul>
  9. 9. timTrack variance-covariance matrix <ul><li>PREVIOUS VERSIONS </li></ul><ul><li>timTrack v1.0 (LAPACK) </li></ul><ul><li>timTrack v1.1 (IPP) </li></ul><ul><li>NEW algebra VERSION </li></ul><ul><li>timTrack v2.0 (LAPACK) </li></ul>
  10. 10. Example Implemented Z Y T1 T2 X
  11. 11. Times for 1.000.000 particles Old Python and Matlab versions (only 500.000 particles) 165m 47.137 s timTrack v2.0 LAPACK 23.615 s timTrack v1.1 intel®IPP 23.495 s timTrack v1.0 LAPACK 31.188 s :)
  12. 12. Next Steps <ul><li>Analyze systematic computing errors </li></ul><ul><li>Check single-precision version </li></ul><ul><li>Parallelize </li></ul><ul><ul><li>Shared memory (OpenMP) </li></ul></ul><ul><ul><li>MPI (master-slave) </li></ul></ul><ul><ul><li>Full distributed </li></ul></ul><ul><li>Implement in GPU </li></ul><ul><li>Study full problem </li></ul>
  13. 13. timTrack v2.1 Next step ( still in progress… ) Parallelims with Intel® MPI libraries Shared parallelism with OpenMP for Multi-core
  14. 14. Future ! <ul><li>timtrack v 3.0 </li></ul><ul><li>CUDA parallel computing architecture in GPUs </li></ul><ul><li>CUDA has several advantages over traditional general purpose computation on GPUs </li></ul><ul><li>* Scattered reads </li></ul><ul><li>* Shared memory </li></ul><ul><li>* Faster downloads from the GPU </li></ul><ul><li>* Full support for integer and bitwise operations </li></ul>
  1. A particular slide catching your eye?

    Clipping is a handy way to collect important slides you want to go back to later.

×