Skip to content
This repository has been archived by the owner on Feb 7, 2024. It is now read-only.

Commit

Permalink
Change legend label from chans to num of chunks in benchmarks.
Browse files Browse the repository at this point in the history
  • Loading branch information
Jan-Willem committed Sep 2, 2021
1 parent e620917 commit 2d39daf
Show file tree
Hide file tree
Showing 14 changed files with 35 additions and 53 deletions.
Binary file modified docs/_media/X1a36/cluster_A001_X133d_X1a36.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
69 changes: 30 additions & 39 deletions docs/_media/X1a36/combine_imgs.txt
Original file line number Diff line number Diff line change
@@ -1,41 +1,32 @@
{\rtf1\ansi\ansicpg1252\cocoartf2513
\cocoatextscaling0\cocoaplatform0{\fonttbl\f0\fswiss\fcharset0 Helvetica;}
{\colortbl;\red255\green255\blue255;}
{\*\expandedcolortbl;;}
\margl1440\margr1440\vieww27180\viewh8400\viewkind0
\pard\tx720\tx1440\tx2160\tx2880\tx3600\tx4320\tx5040\tx5760\tx6480\tx7200\tx7920\tx8640\pardirnatural\partightenfactor0

\f0\fs28 \cf0 \
convert \\( node_A001_X133d_X1a36.png cluster_A001_X133d_X1a36.png -append \\) \\\
\\( spu_node_A001_X133d_X1a36.png spu_cluster_A001_X133d_X1a36.png -append \\) +append x_combined_X1a36.png\
\
convert aws_A001_X133d_X1a36.png spu_awsA001_X133d_X1a36.png +append aws_combined_X1a36.png\
\
convert x_combined_X1a36.png aws_combined_X1a36.png -append combined_X1a36.png\
\

convert \( node_A001_X133d_X1a36.png cluster_A001_X133d_X1a36.png -append \) \
\( spu_node_A001_X133d_X1a36.png spu_cluster_A001_X133d_X1a36.png -append \) +append x_combined_X1a36.png

convert aws_A001_X133d_X1a36.png spu_awsA001_X133d_X1a36.png +append aws_combined_X1a36.png

convert x_combined_X1a36.png aws_combined_X1a36.png -append combined_X1a36.png

###################\
convert \\( node_A001_X1273_X2e3.png cluster_A001_X1273_X2e3.png -append \\) \\\
\\( spu_node_A001_X1273_X2e3.png spu_cluster_A001_X1273_X2e3.png -append \\) +append x_combined_X2e3.png\
\
convert aws_A001_X1273_X2e3.png spu_awsA001_X1273_X2e3.png +append aws_combined_X2e3.png\
\
convert x_combined_X2e3.png aws_combined_X2e3.png -append combined_X2e3.png\
\
\
convert \( node_A001_X1273_X2e3.png cluster_A001_X1273_X2e3.png -append \) \
\( spu_node_A001_X1273_X2e3.png spu_cluster_A001_X1273_X2e3.png -append \) +append x_combined_X2e3.png

convert aws_A001_X1273_X2e3.png spu_awsA001_X1273_X2e3.png +append aws_combined_X2e3.png

convert x_combined_X2e3.png aws_combined_X2e3.png -append combined_X2e3.png


###################\
convert \\( node_A001_X12a3_X3be.png cluster_A001_X12a3_X3be.png -append \\) \\\
\\( spu_node_A001_X12a3_X3be.png spu_cluster_A001_X12a3_X3be.png -append \\) +append x_combined_X3be.png\
\
convert aws_A001_X12a3_X3be.png spu_awsA001_X12a3_X3be.png +append aws_combined_X3be.png\
\
convert x_combined_X3be.png aws_combined_X3be.png -append combined_X3be.png\
\
###############\
\
\
convert \\( node_A001_X1273_Xc66.png cluster_A001_X1273_Xc66.png -append \\) \\\
\\( spu_node_A001_X1273_Xc66.png spu_cluster_A001_X1273_Xc66.png -append \\) +append x_combined_Xc66.png\
\
convert aws_A001_X1273_Xc66.png spu_awsA001_X1273_Xc66.png +append aws_combined_Xc66.png\
\
convert x_combined_Xc66.png aws_combined_Xc66.png -append combined_Xc66.png}
convert \( node_A001_X12a3_X3be.png cluster_A001_X12a3_X3be.png -append \) \
\( spu_node_A001_X12a3_X3be.png spu_cluster_A001_X12a3_X3be.png -append \) +append x_combined_X3be.png

convert aws_A001_X12a3_X3be.png spu_awsA001_X12a3_X3be.png +append aws_combined_X3be.png

convert x_combined_X3be.png aws_combined_X3be.png -append combined_X3be.png

###############
convert \( node_A001_X1273_Xc66.png cluster_A001_X1273_Xc66.png -append \) \
\( spu_node_A001_X1273_Xc66.png spu_cluster_A001_X1273_Xc66.png -append \) +append x_combined_Xc66.png

convert aws_A001_X1273_Xc66.png spu_awsA001_X1273_Xc66.png +append aws_combined_Xc66.png

convert x_combined_Xc66.png aws_combined_Xc66.png -append combined_Xc66.png
Binary file modified docs/_media/X1a36/combined_X1a36.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/X1a36/node_A001_X133d_X1a36.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/X2e3/cluster_A001_X1273_X2e3.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/X2e3/combined_X2e3.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/X2e3/node_A001_X1273_X2e3.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/X3be/cluster_A001_X12a3_X3be.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/X3be/combined_X3be.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/X3be/node_A001_X12a3_X3be.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/Xc66/cluster_A001_X1273_Xc66.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/Xc66/combined_Xc66.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file modified docs/_media/Xc66/node_A001_X1273_Xc66.png
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
19 changes: 5 additions & 14 deletions docs/benchmarking.ipynb
Original file line number Diff line number Diff line change
Expand Up @@ -120,15 +120,7 @@
"- Image size (x,y,chan,pol): (1000, 1000, 3853, 2)\n",
"- Data Volume (vis.zarr and img.zarr): 304 GB\n",
"\n",
"![im14](https://raw.githubusercontent.com/casangi/cngi_prototype/master/docs/_media/Xc66/combined_Xc66.png)\n",
"\n",
"\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"![im14](https://raw.githubusercontent.com/casangi/cngi_prototype/master/docs/_media/X3be/combined_X3be.png)\n",
"\n",
"\n"
]
Expand All @@ -150,18 +142,17 @@
"source": [
"**Single Machine**\n",
"\n",
"The total runtime of the prototype mosaic gridder was less than the 6.1 and 6.2 reference implementations in most cases. The prototype standard gridder has comparable performance for all but the least-optimal chunk size selection. \n",
"The total runtime of the prototype has comparable performance to the CASA 6.2 reference implementations for all datasets.\n",
"\n",
"There does not appear to be a performance penalty associated with the adoption of a pure Python framework in comparison to the compiled C++/Fortran reference implementation. This is likely due in large part to the prototype's reliance on the `numba` Just-In-Time (JIT) transpiler and the C foreign function interface relied on by third-party framework packages including `numpy` and `scipy`.\n",
"\n",
"The Fortran gridding code in CASA appears slightly more efficient than the JIT-decorated Python code in the prototype. However, the test implementation more efficiently handles chunked data and does not have intermediate steps where data is written to disk, whereas CASA generates TempLattice files to store intermediate files.\n",
"The Fortran gridding code in CASA is slightly more efficient than the JIT-decorated Python code in the prototype. However, the test implementation more efficiently handles chunked data and does not have intermediate steps where data is written to disk, whereas CASA generates TempLattice files to store intermediate files.\n",
"\n",
"**Multi-Node**\n",
"\n",
"The total runtime of the prototype mosaic and standard gridders was less than the 6.1 and 6.2 reference implementations in all cases. \n",
"\n",
"There does not appear to be a performance penalty associated with the adoption of a pure Python framework for distributed scheduling in comparison to the MPI-based reference implementation. This is likely due in part to the graph optimization of the task scheduler, which includes overhead that begins to dominate the total runtime at higher levels of concurrency.\n",
"The total runtime of the prototype mosaic and standard gridders was less than the 6.2 reference implementations except for a couple of suboptimal chunking sizes. For the two larger datasets 2017.1.00717.S and 2017.1.00983.S a significant speedup of up to six times and ten times respectively, was achieved. Furthermore, the performance of CASA 6.2 stagnates after a single node, because the output images can not be written to disk in parallel. This is due to the CASACORE table system that does not support parallel writes.\n",
"\n",
"**Chunking**\n",
"\n",
"**Comparison of CASA versions**\n",
"\n",
Expand Down

0 comments on commit 2d39daf

Please sign in to comment.