diff options
| author | Erik Schnetter <schnetter@cct.lsu.edu> | 2009-09-03 16:19:15 -0500 |
|---|---|---|
| committer | Barry Wardell <barry.wardell@gmail.com> | 2011-12-14 16:42:31 +0000 |
| commit | 11c4d98017cbb86d08e15fd1b549180184b58a26 (patch) | |
| tree | 2546a154c6f7bc0bec87de7316125ae7d1453569 /Carpet/CarpetWeb/olds.html | |
| parent | f520477b1c14e02f1495cfa8d3e09f4e21ab34d0 (diff) | |
Import Carpet
Ignore-this: 309b4dd613f4af2b84aa5d6743fdb6b3
Diffstat (limited to 'Carpet/CarpetWeb/olds.html')
| -rw-r--r-- | Carpet/CarpetWeb/olds.html | 215 |
1 files changed, 213 insertions, 2 deletions
diff --git a/Carpet/CarpetWeb/olds.html b/Carpet/CarpetWeb/olds.html index 8bd58f747..1d29d2cc8 100644 --- a/Carpet/CarpetWeb/olds.html +++ b/Carpet/CarpetWeb/olds.html @@ -1,10 +1,10 @@ -<?xml version="1.0" encoding="ISO-8859-15"?> +<?xml version="1.0" encoding="utf-8"?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en"> <head> - <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-15" /> + <meta http-equiv="Content-Type" content="text/html; charset=utf-8" /> <title>Old News</title> </head> @@ -14,6 +14,217 @@ <p><a href="index.html"><b>New News...</b></a></p> + <table><tr><td valign="top"> + <p><b>March 1, 2008:</b> Carpet has a logo! This logo is + a <a href="http://en.wikipedia.org/wiki/Sierpinski_carpet">Sierpiński + carpet</a>, which is a fractal pattern with + a <a href="http://en.wikipedia.org/wiki/Hausdorff_dimension">Hausdorff + dimension</a> of 1.89279.</p> + </td><td valign="top"> + <p><a href="logo/Sierpinski.pdf"><img src="logo/Sierpinski.png" + width="100" alt="Carpet logo (a Sierpiński + carpet)" /></a></p> + </td></tr></table> + + <p><b>March 1, 2008:</b> We have improved the development version + of Carpet significantly:<br /></p> + <ul> + <li><p>The data structures and algorithms storing and handling + the communication schedule are much more efficient on large + numbers (several hundred or more) processors. This makes Carpet + scale to more than 8,000 cores.</p></li> + + <li><p>The interface for defining and making dynamic changes to + grid hierarchies is simpler, and buffer zones are handled in a + cleaner manner. This makes it easier to write user code which + defines or updates the grid hierarchy, and reduces the chance of + inconsistencies therein.</p></li> + + <li><p>During checkpointing and recovery, the grid structure is + saved and restored by default. This avoids accidental changes + upon recovery.</p></li> + + <li><p>The efficiency of I/O has been increased, especially for + HDF5 based binary I/O. It is possible to combine several + variables into one file to reduce the number of output + files.</p></li> + + <li><p>A new thorn LoopControl offers iterators over grid + points, implemented as C-style macros. These iterators allow + additional important loop-level optimisations, such + as <a href="http://en.wikipedia.org/wiki/Loop_tiling">loop + tiling</a> or + <a href="http://www.openmp.org/">OpenMP</a> parallelisation. + Efficient cache handling and hybrid communication models have a + large potential for performance improvements on current and + future architectures.</p></li> + </ul> + <p>More details can be found <a href="version-4.html">here</a>. + These improvements are largely due + to <a href="http://www.cct.lsu.edu/~eschnett/">Erik Schnetter</a> + (LSU), + <a href="http://www.aei.mpg.de/~tradke/">Thomas Radke</a> (AEI), and + <a href="http://www.tapir.caltech.edu/~cott/">Christian D. Ott</a> + (UA). Special thanks go to Christian Reisswig and Luca + Baiotti.</p> + + <p><b>March 1, 2008:</b> The development version of Carpet is now + maintained using <a href="http://git.or.cz/">git</a> instead + of <a href="http://www.darcs.net/">darcs</a>. Git offers a very + similar set of features to darcs, most importantly supporting + decentralised development. Git has a much larger user community + than darcs, and we hope that this makes it easier to use. + The <a href="get-carpet.html">download instructions</a> contain + details on using git to obtain Carpet, and point to further + information. (The darcs repository for the development version + will not see any further changes.)</p> + + <p><b>March 1, 2008:</b> The repository for the development + version of Carpet moved today to + a <a href="http://carpetcode.dyndns.org/">new server</a>. The + stable versions of Carpet continue to be served from the old + server for the time being. We plan to move all of carpetcode.org + to this new server in the future. The new server is a courtesy + of <a href="http://www.tapir.caltech.edu/~cott/">Christian + D. Ott</a>.</p> + + <table><tr><td valign="top"> + <p><b>January 14, 2008:</b> Carpet's communication + infrastructure has been improved significantly, making Carpet + scale to at least 4,000 processors, including mesh refinement. + Using "friendly user time" + on <a + href="http://www.tacc.utexas.edu/services/userguides/ranger/">Ranger</a>, + the new 60,000 + core <a href="http://www.teragrid.org/">TeraGrid</a> + supercomputer + at <a href="http://www.tacc.utexas.edu/">TACC</a>, we measured + the benchmark results below for a numerical relativity kernel + solving the BSSN equations. These benchmarks emply a hybrid + communication scheme + combining <a href="http://www-unix.mcs.anl.gov/mpi/">MPI</a> + and + <a href="http://www.openmp.org/">OpenMP</a>, using the shared + memory capabilities of Ranger's nodes to reduce the memory + overhead of parallelisation. We are grateful for the help we + received from Ranger's support team.</p> + <p>The graph below shows weak scaling tests for both unigrid and + mesh refinement benchmarks. The problem size per core was + kept fixed, and there were 4 OpenMP threads per MPI process, + with 1 MPI process per socket. The benchmark was also run + with the PUGH driver for comparison for certain core counts. + As the graphs show, this benchmark scales near perfectly for + unigrid, and has only small variations in run time for nine + levels of mesh refinement.</p> + </td><td valign="top"> + <p><a + href="scaling-ranger/results-ranger.pdf"><img + src="scaling-ranger/results-ranger.png" width="234" + alt="Scaling graph for Ranger" /></a></p> + </td></tr></table> + + <hr /> + + <p><b>October 3, 2007:</b> Carpet's timing infrastructure has been + extended to automatically measure both time spent computing and + time spent in I/O. The performance of large simulations depends + not only on the computational efficiency and communication + latency, but also on the throughput to file servers. These new + statistics give a real-time overview and can point out + performance problems. The statistics are collected in the + existing <tt>Carpet::timing</tt> variables.</p> + + <p><b>August 30, 2007:</b> So far this year, ten of the + publications from three research groups examining the dynamics + of binary black hole systems are based on simulations performed + with Cactus and Carpet:<br /> + + <a href="http://www.journals.uchicago.edu/ApJ/journal/issues/ApJ/v661n1/71342/71342.html">Astrophys. J. <b>661</b>, 430-436 (2007)</a> + (<a href="http://arxiv.org/abs/gr-qc/0701143">arXiv:gr-qc/0701143</a>)<br /> + + <a href="http://link.aps.org/abstract/PRL/v99/e041102">Phys. Rev. Lett. <b>99</b>, 041102 (2007)</a> + (<a href="http://arxiv.org/abs/gr-qc/0701163">arXiv:gr-qc/0701163</a>)<br /> + + <a href="http://www.journals.uchicago.edu/ApJ/journal/issues/ApJL/v659n1/21515/brief/21515.abstract.html">Astrophys. J. <b>659</b>, L5-L8 (2007)</a> + (<a href="http://arxiv.org/abs/gr-qc/0701164">arXiv:gr-qc/0701164</a>)<br /> + + <a href="http://link.aps.org/abstract/PRL/v98/e231102">Phys. Rev. Lett. <b>98</b>, 231102 (2007)</a> + (<a href="http://arxiv.org/abs/gr-qc/0702133">arXiv:gr-qc/0702133</a>)<br /> + + <a href="http://www.iop.org/EJ/abstract/0264-9381/24/15/009/">Class. Quantum Grav. <b>24</b>, 3911-3918 (2007)</a> + (<a href="http://arxiv.org/abs/gr-qc/0701038">arXiv:gr-qc/0701038</a>)<br /> + + <a href="http://arxiv.org/abs/0705.3829">arXiv:0705.3829 [gr-qc]</a><br /> + + <a href="http://arxiv.org/abs/0706.2541">arXiv:0706.2541 [gr-qc]</a><br /> + + <a href="http://arxiv.org/abs/0707.2559">arXiv:0707.2559 [gr-qc]</a><br /> + + <a href="http://arxiv.org/abs/0708.3999">arXiv:0708.3999 [gr-qc]</a><br /> + + <a href="http://arxiv.org/abs/0708.4048">arXiv:0708.4048 [gr-qc]</a><br /> + These publications mainly examine the spin dynamics and the + gravitational wave recoil in BBH systems. Since not all + research groups use Cactus and Carpet, this represents only part + of the published work on this subject.</p> + + <table><tr><td valign="top"> + <p><b>August 26, 2007:</b> In experiments with hybrid + communication schemes + combining <a href="http://www-unix.mcs.anl.gov/mpi/">MPI</a> + and + <a href="http://www.openmp.org/">OpenMP</a>, we found a 20% + speed improvement when using a single node + of <a + href="http://www.ncsa.uiuc.edu/UserInfo/Resources/Hardware/Intel64Cluster/">Abe</a> + at <a href="http://www.ncsa.uiuc.edu">NCSA</a>, and a + substantial scaling improvement when using 1024 and more CPUs. + (Abe has 8 CPUs per node.) These experiments included cache + optimisations when traversing the 3D arrays. The tests were + performed with a modified version of + the <a + href="http://www.cactuscode.org/">Cactus</a> <a + href="http://www.cactuscode.org/WaveToyDemo/">WaveToy</a> + example application without using I/O or analysis methods.</p> + </td><td valign="top"> + <p><a + href="hybrid-scaling/results-wavetoy-abe.pdf"><img + src="hybrid-scaling/results-wavetoy-abe.png" width="200" + alt="Scaling graph for Abe" /></a></p> + </td></tr></table> + + <p><b>August 15, 2007:</b> We are happy to hear that our + proposal <i>ALPACA: Cactus tools for Application Level Profiling + And Correctness Analysis</i> will be funded by + <a + href="http://www.nsf.gov/">NSF's</a> <a + href="http://www.nsf.gov/publications/pub_summ.jsp?ods_key=nsf07503">SDCI</a> + programme for three years. + The <a + href="http://www.cactuscode.org/Development/alpaca">ALPACA</a> + project is aiming at developing complex, collaborative + scientific applications, appropriate for highly scalable + hardware architectures, providing fault tolerance, advanced + debugging, and transparency against new developments in + communication, programming, and execution models. Such tools + are especially rare at the application level, where they are + most critically needed.</p> + + <p><b>July 31, 2007:</b> We are happy to hear that our + proposal <i>XiRel: Cyberinfrastructure for Numerical + Relativity</i> will be funded by + <a href="http://www.nsf.gov/">NSF's</a> <a href="http://www.nsf.gov/funding/pgm_summ.jsp?pims_id=6681">PIF</a> + programme for three + years. <a href="http://www.cct.lsu.edu/xirel/">XiRel</a> is + collaborative proposal + by <a href="http://www.cct.lsu.edu/">LSU</a>, <a href="http://gravity.psu.edu/numrel/">PSU</a>, + and <a href="http://www.phys.utb.edu/numrel/">UTB</a> + (now <a href="http://ccrg.rit.edu/">RIT</a>). The central goal of + XiRel is the development of a highly scalable, efficient, and + accurate adaptive mesh refinement layer based on the current + Carpet driver, which will be fully integrated and supported in + Cactus and optimised for numerical relativity.</p> + <p><b>February 26, 2007:</b> The thorn <tt>LSUPETSc</tt> implements a generic elliptic solver for Carpet's multi-patch infrastructure, based |