forgotten due to .gitignore

.gitignore

@@ -1,11 +0,0 @@
-# Ignore everything
-*
-
-# But not these!
-!.gitignore
-!README.md
-!*.py
-!*.template
-# Optional: Keep subdirectories and their Python files
-
-!*/

results-and-plotting/python/notebooks/bcast_analysis.ipynb

@@ -0,0 +1,175 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "id": "da7c16b4",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "import matplotlib.pyplot as plt\n",
+    "import seaborn as sns\n",
+    "from scipy.optimize import curve_fit\n",
+    "from matplotlib.cm import get_cmap"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "47341b1d",
+   "metadata": {},
+   "source": [
+    "# Alltoall "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "1cc39aab",
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "",
+     "evalue": "",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31mnotebook controller is DISPOSED. \n",
+      "\u001b[1;31mView Jupyter <a href='command:jupyter.viewOutput'>log</a> for further details."
+     ]
+    }
+   ],
+   "source": [
+    "df_multinode = pd.read_csv(\"../data/data-multi-defand100cflag.csv\",delimiter = \",\")\n",
+    "df_multinode['benchmark_type'].unique()\n",
+    "df_gather = df_multinode[df_multinode[\"benchmark_type\"]==\"Bcast\"][df_multinode['msg_size_bytes']>1024][df_multinode['off_cache_flag']==-1]\n",
+    "df_gather.columns.tolist()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "4336d3c6",
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "",
+     "evalue": "",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31mnotebook controller is DISPOSED. \n",
+      "\u001b[1;31mView Jupyter <a href='command:jupyter.viewOutput'>log</a> for further details."
+     ]
+    }
+   ],
+   "source": [
+    "def model(proc_num, alpha, beta, msg_size):\n",
+    "    return (alpha * msg_size * (proc_num - 72) * 72) / (12.5 * 1e3) + 1e6*beta\n",
+    "\n",
+    "results = []\n",
+    "msg_sizes = sorted(df_gather['msg_size_bytes'].unique())\n",
+    "n_rows = int(np.ceil(len(msg_sizes) / 3))\n",
+    "n_cols = min(len(msg_sizes), 3)\n",
+    "fig, axes = plt.subplots(n_rows, n_cols, figsize=(5*n_cols, 4*n_rows), squeeze=False)\n",
+    "cmap = get_cmap('tab10')\n",
+    "\n",
+    "for idx, (msg_size, group) in enumerate(df_gather.groupby('msg_size_bytes')):\n",
+    "    x = group['proc_num'].values.copy()\n",
+    "    y = group['t_avg_usec'].values.copy()\n",
+    "    sorted_indices = np.argsort(x)\n",
+    "    x = x[sorted_indices]\n",
+    "    y = y[sorted_indices]\n",
+    "    fit_func = lambda proc_num, alpha, beta: model(proc_num, alpha, beta, msg_size)\n",
+    "    popt, _ = curve_fit(fit_func, x, y, bounds=([1, 0], [np.inf, np.inf]))\n",
+    "    alpha, beta = popt\n",
+    "    results.append({'msg_size_bytes': msg_size, 'alpha': alpha, 'beta': beta})\n",
+    "\n",
+    "    x_fit = np.linspace(min(x), max(x), 100)\n",
+    "    y_fit = fit_func(x_fit, alpha, beta)\n",
+    "    y_speed = model(x_fit,1,0,msg_size)\n",
+    "    row, col = divmod(idx, n_cols)\n",
+    "    ax = axes[row][col]\n",
+    "\n",
+    "    color = cmap(idx % 10)\n",
+    "    # ax.scatter(x, y/1e6, label='Data', color=color)\n",
+    "    ax.plot(x, y/1e6, label='Data', color=color)\n",
+    "    # ax.plot(x_fit, y_fit/1e6, linestyle='--', color=color, alpha=0.5, label='Fit')\n",
+    "    # ax.plot(x_fit, y_speed/1e6, linestyle='--', color='red', alpha=0.1, label='Fit')\n",
+    "    ax.set_title(f'msg_size: {msg_size} bytes')\n",
+    "    ax.set_xlabel('num. proc.')\n",
+    "    ax.set_ylabel('Average Time [s]')\n",
+    "    ax.set_xticks(x)\n",
+    "    ax.grid(True)\n",
+    "    max_data =(x[-1]-72)*72*msg_size\n",
+    "    min_data =(x[0]-72)*72*msg_size\n",
+    "\n",
+    "    textstr = \"\"\n",
+    "    # if(max_data > 1e9):\n",
+    "    #     textstr+=f\"max data = {max_data/1e9:0.2f}GB\\n\" \n",
+    "    # else:\n",
+    "    #     textstr+=f\"max data = {max_data/1e6:0.2f}MB\\n\" \n",
+    "\n",
+    "    # if(min_data > 1e9):\n",
+    "    #     textstr+=f\"min data = {min_data/1e9:0.2f}GB\\n\" \n",
+    "    # else:\n",
+    "    #     textstr+=f\"min data = {min_data/1e6:0.2f}MB\\n\" \n",
+    "    # textstr += r\"$\\alpha$\" +f\"= {alpha:.3e}\\n\"+r\"$b_{eff}=$\"+f\"{12.5/alpha:0.3f}Gbps\\n\"+\\\n",
+    "    #            r\"$\\beta$\"+f\"= {beta:.3e} s\"\n",
+    "    # ax.text(0.95, 0.05, textstr, transform=ax.transAxes,\n",
+    "    #         fontsize=10, verticalalignment='bottom',\n",
+    "    #         horizontalalignment='right',\n",
+    "    #         bbox=dict(boxstyle='round', facecolor='white', alpha=0.8))\n",
+    "\n",
+    "fig.suptitle('Alltoall Time Fit per Message Size\\nDots = Data Points | Dashed Lines = Fits\\n off_mem=-1', fontsize=14)\n",
+    "fig.tight_layout(rect=[0, 0.03, 1, 0.95])\n",
+    "# plt.savefig(\"plots/alltoall_analysis.png\",dpi=300)\n",
+    "plt.show()\n",
+    "\n",
+    "fit_results = pd.DataFrame(results)\n",
+    "fit_results['inv_alpha'] = 1 / fit_results['alpha']\n",
+    "print(fit_results)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "ce632d6f",
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "",
+     "evalue": "",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31mnotebook controller is DISPOSED. \n",
+      "\u001b[1;31mView Jupyter <a href='command:jupyter.viewOutput'>log</a> for further details."
+     ]
+    }
+   ],
+   "source": [
+    "df_gather[df_gather['msg_size_bytes']==1048576]"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "data",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.13.2"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

results-and-plotting/python/scripts/plot_all.py

@@ -0,0 +1,69 @@
+import pandas as pd
+import numpy as np
+import matplotlib.pyplot as plt
+import seaborn as sns
+from scipy.optimize import curve_fit
+import matplotlib.cm as cm
+
+
+def max_transfer_size(msg_size, np_procs, benchmark_type):
+    if benchmark_type == 'Allgather':
+        return (np_procs-72)*msg_size
+    elif benchmark_type == 'Scatter':
+        return (np_procs-72)*msg_size  # ?
+    elif benchmark_type == 'Alltoall':
+        return 72*(np_procs-72)*msg_size
+    elif benchmark_type == 'Bcast':
+        return msg_size
+    elif benchmark_type == 'Gather':
+        return (np_procs)*msg_size  # ?
+    elif benchmark_type == 'Reduce_scatter':
+        return 0.25*(np_procs-72)*(1/72)*msg_size  # ?
+    elif benchmark_type == 'Allreduce':
+        return 0.25*(np_procs-72)*(1/72)*msg_size
+    elif benchmark_type == 'Reduce':
+        return 0.25*(np_procs-72)*(1/72)*msg_size
+
+
+data_file = "data/data-multi-defand100cflag.csv"
+df_multinode = pd.read_csv(data_file, delimiter=',')
+df_multinode_offdef = df_multinode[df_multinode['off_cache_flag'] == 100]
+benchmarks = df_multinode_offdef['benchmark_type'].unique().tolist()
+benchmarks = [x for x in benchmarks if x[-1] != 'v']
+print(benchmarks)
+df_multinode_offdef = df_multinode_offdef[df_multinode_offdef['benchmark_type'].isin(
+    benchmarks)][df_multinode_offdef['msg_size_bytes'] > 1000]
+
+df_multinode_offdef["max_transfer"] = df_multinode_offdef.apply(
+    lambda row: max_transfer_size(
+        msg_size=row["msg_size_bytes"],
+        np_procs=row["proc_num"],
+        benchmark_type=row["benchmark_type"]
+    ),
+    axis=1
+)
+
+df_multinode_offdef["bytes/usec"] = df_multinode_offdef["max_transfer"] / \
+    df_multinode_offdef["t_avg_usec"]
+df_multinode_offdef = df_multinode_offdef[df_multinode_offdef['benchmark_type']!='Allgather'][df_multinode_offdef['benchmark_type']!='Alltoall']
+df_multinode_offdef = df_multinode_offdef[['benchmark_type','msg_size_bytes','t_avg_usec','proc_num']]
+
+plt.figure(figsize=(16, 9))
+sns.barplot(
+    data=df_multinode_offdef,
+    x="benchmark_type",
+    y="t_avg_usec",
+    dodge=True,
+    hue=df_multinode_offdef["msg_size_bytes"].astype(str),
+)
+# plt.yscale("log")
+plt.title("Average Time (usec) per Benchmark Type and Message Size")
+plt.ylabel("Average Time (usec)")
+plt.xlabel("Benchmark Type")
+plt.xticks(rotation=45)
+plt.legend(title="Message Size (bytes)")
+plt.tight_layout()
+# plt.show()
+plt.savefig("./plots/benchmark_avg_time_barplot.png", dpi=300)
+plt.close()
+

results-and-plotting/python/scripts/plot_fast_group.py

@@ -0,0 +1,64 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+import numpy as np
+from mpl_toolkits.mplot3d import Axes3D
+
+data_file = "data/data-multi-defand100cflag.csv"
+df_multinode = pd.read_csv(data_file, delimiter=',')
+df_multinode_offdef = df_multinode[df_multinode['off_cache_flag'] == 100]
+df_multinode_offdef = df_multinode_offdef[['benchmark_type','msg_size_bytes','t_avg_usec','proc_num']]
+
+benchmarks = df_multinode_offdef['benchmark_type'].unique().tolist()
+benchmarks = [x for x in benchmarks if x[-1] != 'v']
+df_multinode_offdef = df_multinode_offdef[df_multinode_offdef['benchmark_type'].isin(
+    benchmarks)][df_multinode_offdef['msg_size_bytes'] > 1000]
+fast_benchmarks = ["Allreduce","Bcast","Reduce","Reduce_scatter"]
+df_multinode_offdef = df_multinode_offdef[df_multinode_offdef["benchmark_type"].isin(fast_benchmarks)]
+
+plt.figure(figsize=(16, 9))
+sns.barplot(
+    data=df_multinode_offdef,
+    x="benchmark_type",
+    y="t_avg_usec",
+    dodge=True,
+    hue=df_multinode_offdef["msg_size_bytes"].astype(str),
+)
+
+plt.ylim(0)
+plt.title("Average Time (usec) per Benchmark Type and Message Size")
+plt.ylabel("Average Time (usec)")
+plt.xlabel("Benchmark Type")
+plt.xticks(rotation=45)
+plt.legend(title="Message Size (bytes)")
+plt.tight_layout()
+plt.savefig("./plots/fbenchmarks_avg_time_barplot.png", dpi=300)
+plt.close()
+
+df_allreduce= df_multinode_offdef[df_multinode_offdef["benchmark_type"]=="Allreduce"]
+df_allreduce = df_allreduce[['msg_size_bytes','t_avg_usec','proc_num']]
+df_allreduce = df_allreduce[df_allreduce['msg_size_bytes']>2**17]
+pivot = df_allreduce.pivot(index="msg_size_bytes", columns="proc_num", values="t_avg_usec")
+X = pivot.columns.values   # proc_num
+Y = pivot.index.values     # msg_size_bytes
+X, Y = np.meshgrid(X, Y)
+Z = pivot.values
+
+fig = plt.figure(figsize=(16, 9))
+ax = fig.add_subplot(111, projection='3d')
+surf = ax.plot_surface(X, Y, Z, cmap="viridis", edgecolor='k')
+cbar = fig.colorbar(surf, ax=ax, shrink=0.6, pad=0.01, location='left')
+cbar.set_label("Average Time (μs)")
+ax.set_xlabel("Process Count")
+ax.set_ylabel("Message Size (B)")
+ax.set_zlabel("Average Time (μs)")
+ax.set_title("Allreduce")
+ax.set_xticks(pivot.columns.values)           # use the actual process count values
+ax.set_xticklabels(pivot.columns.values)
+ax.set_yticks(Y[:, 0])
+ymin, ymax = ax.get_ylim()
+ax.set_ylim(ymin*0.8, ymax)   # 30% more space at top
+ax.set_yticklabels([f"$2^{{{int(np.log2(v))}}}$" for v in Y[:, 0]])
+plt.tight_layout()
+plt.savefig("./plots/allreduce_surface.png", dpi=300)
+plt.close()

results-and-plotting/python/scripts/plot_mid_group.py

@@ -0,0 +1,66 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+import numpy as np
+from mpl_toolkits.mplot3d import Axes3D
+
+data_file = "data/data-multi-defand100cflag.csv"
+df_multinode = pd.read_csv(data_file, delimiter=',')
+df_multinode_offdef = df_multinode[df_multinode['off_cache_flag'] == 100]
+df_multinode_offdef = df_multinode_offdef[['benchmark_type','msg_size_bytes','t_avg_usec','proc_num']]
+
+benchmarks = df_multinode_offdef['benchmark_type'].unique().tolist()
+benchmarks = [x for x in benchmarks if x[-1] != 'v']
+df_multinode_offdef = df_multinode_offdef[df_multinode_offdef['benchmark_type'].isin(
+    benchmarks)][df_multinode_offdef['msg_size_bytes'] > 1000]
+# fast_benchmarks = ["Allreduce","Bcast","Reduce","Reduce_scatter"]
+
+medium_benchmarks = ["Gather","Scatter"]
+df_multinode_offdef = df_multinode_offdef[df_multinode_offdef["benchmark_type"].isin(medium_benchmarks)]
+
+plt.figure(figsize=(16, 9))
+sns.barplot(
+    data=df_multinode_offdef,
+    x="benchmark_type",
+    y="t_avg_usec",
+    dodge=True,
+    hue=df_multinode_offdef["msg_size_bytes"].astype(str),
+)
+
+plt.ylim(0)
+plt.title("Average Time (usec) per Benchmark Type and Message Size")
+plt.ylabel("Average Time (usec)")
+plt.xlabel("Benchmark Type")
+plt.xticks(rotation=45)
+plt.legend(title="Message Size (bytes)")
+plt.tight_layout()
+plt.savefig("./plots/mbenchmarks_avg_time_barplot.png", dpi=300)
+plt.close()
+
+df_gather = df_multinode_offdef[df_multinode_offdef['benchmark_type']=='Gather']
+df_gather = df_gather[['msg_size_bytes','t_avg_usec','proc_num']]
+df_gather = df_gather[df_gather['msg_size_bytes']>2**17]
+pivot = df_gather.pivot(index="msg_size_bytes", columns="proc_num", values="t_avg_usec")
+X = pivot.columns.values   # proc_num
+Y = pivot.index.values     # msg_size_bytes
+X, Y = np.meshgrid(X, Y)
+Z = pivot.values
+
+fig = plt.figure(figsize=(16, 9))
+ax = fig.add_subplot(111, projection='3d')
+surf = ax.plot_surface(X, Y, Z, cmap="viridis", edgecolor='k')
+cbar = fig.colorbar(surf, ax=ax, shrink=0.6, pad=0.01, location='left')
+cbar.set_label("Average Time (μs)")
+ax.set_xlabel("Process Count")
+ax.set_ylabel("Message Size (B)")
+ax.set_zlabel("Average Time (μs)")
+ax.set_title("Gather")
+ax.set_xticks(pivot.columns.values)           # use the actual process count values
+ax.set_xticklabels(pivot.columns.values)
+ax.set_yticks(Y[:, 0])
+ymin, ymax = ax.get_ylim()
+ax.set_ylim(ymin*0.8, ymax)   # 30% more space at top
+ax.set_yticklabels([f"$2^{{{int(np.log2(v))}}}$" for v in Y[:, 0]])
+plt.tight_layout()
+plt.savefig("./plots/gather_surface.png", dpi=300)
+plt.close()

results-and-plotting/python/scripts/plot_slow_group.py

@@ -0,0 +1,93 @@
+import pandas as pd
+import matplotlib.pyplot as plt
+import seaborn as sns
+import numpy as np
+from mpl_toolkits.mplot3d import Axes3D
+
+data_file = "data/data-multi-defand100cflag.csv"
+df_multinode = pd.read_csv(data_file, delimiter=',')
+df_multinode_offdef = df_multinode[df_multinode['off_cache_flag'] == 100]
+df_multinode_offdef = df_multinode_offdef[['benchmark_type','msg_size_bytes','t_avg_usec','proc_num']]
+
+benchmarks = df_multinode_offdef['benchmark_type'].unique().tolist()
+benchmarks = [x for x in benchmarks if x[-1] != 'v']
+df_multinode_offdef = df_multinode_offdef[df_multinode_offdef['benchmark_type'].isin(
+    benchmarks)][df_multinode_offdef['msg_size_bytes'] > 1000]
+
+slow_benchmarks = ["Alltoall","Allgather"]
+df_multinode_offdef = df_multinode_offdef[df_multinode_offdef["benchmark_type"].isin(slow_benchmarks)]
+
+plt.figure(figsize=(16, 9))
+sns.barplot(
+    data=df_multinode_offdef,
+    x="benchmark_type",
+    y="t_avg_usec",
+    dodge=True,
+    hue=df_multinode_offdef["msg_size_bytes"].astype(str),
+)
+
+plt.ylim(0)
+plt.title("Average Time (usec) per Benchmark Type and Message Size")
+plt.ylabel("Average Time (usec)")
+plt.xlabel("Benchmark Type")
+plt.xticks(rotation=45)
+plt.legend(title="Message Size (bytes)")
+plt.tight_layout()
+plt.savefig("./plots/sbenchmarks_avg_time_barplot.png", dpi=300)
+plt.close()
+
+df_alltoall = df_multinode_offdef[df_multinode_offdef['benchmark_type']=='Alltoall']
+df_alltoall = df_alltoall[['msg_size_bytes','t_avg_usec','proc_num']]
+df_alltoall = df_alltoall[df_alltoall['msg_size_bytes']>2**17]
+pivot = df_alltoall.pivot(index="msg_size_bytes", columns="proc_num", values="t_avg_usec")
+X = pivot.columns.values   # proc_num
+Y = pivot.index.values     # msg_size_bytes
+X, Y = np.meshgrid(X, Y)
+Z = pivot.values
+
+fig = plt.figure(figsize=(16, 9))
+ax = fig.add_subplot(111, projection='3d')
+surf = ax.plot_surface(X, Y, Z, cmap="viridis", edgecolor='k')
+cbar = fig.colorbar(surf, ax=ax, shrink=0.6, pad=0.01, location='left')
+cbar.set_label("Average Time (μs)")
+ax.set_xlabel("Process Count")
+ax.set_ylabel("Message Size (B)")
+ax.set_zlabel("Average Time (μs)")
+ax.set_title("Alltoall")
+ax.set_xticks(pivot.columns.values)           # use the actual process count values
+ax.set_xticklabels(pivot.columns.values)
+ax.set_yticks(Y[:, 0])
+ymin, ymax = ax.get_ylim()
+ax.set_ylim(ymin*0.8, ymax)   # 30% more space at top
+ax.set_yticklabels([f"$2^{{{int(np.log2(v))}}}$" for v in Y[:, 0]])
+plt.tight_layout()
+plt.savefig("./plots/alltoall_surface.png", dpi=300)
+plt.close()
+
+df_allgather = df_multinode_offdef[df_multinode_offdef['benchmark_type']=='Allgather']
+df_allgather = df_allgather[['msg_size_bytes','t_avg_usec','proc_num']]
+df_allgather = df_allgather[df_allgather['msg_size_bytes']>2**17]
+pivot = df_allgather.pivot(index="msg_size_bytes", columns="proc_num", values="t_avg_usec")
+X = pivot.columns.values   # proc_num
+Y = pivot.index.values     # msg_size_bytes
+X, Y = np.meshgrid(X, Y)
+Z = pivot.values
+
+fig = plt.figure(figsize=(16, 9))
+ax = fig.add_subplot(111, projection='3d')
+surf = ax.plot_surface(X, Y, Z, cmap="viridis", edgecolor='k')
+cbar = fig.colorbar(surf, ax=ax, shrink=0.6, pad=0.01, location='left')
+cbar.set_label("Average Time (μs)")
+ax.set_xlabel("Process Count")
+ax.set_ylabel("Message Size (B)")
+ax.set_zlabel("Average Time (μs)")
+ax.set_title("Allgather")
+ax.set_xticks(pivot.columns.values)           # use the actual process count values
+ax.set_xticklabels(pivot.columns.values)
+ax.set_yticks(Y[:, 0])
+ymin, ymax = ax.get_ylim()
+ax.set_ylim(ymin*0.8, ymax)   # 30% more space at top
+ax.set_yticklabels([f"$2^{{{int(np.log2(v))}}}$" for v in Y[:, 0]])
+plt.tight_layout()
+plt.savefig("./plots/allgather_surface.png", dpi=300)
+plt.close()

templates/multinode.template

@@ -3,13 +3,36 @@
 #SBATCH --output={output_dir}{job_name}_{n_procs}.out
 #SBATCH --error={err_dir}{job_name}_{n_procs}.err
 #SBATCH --nodes={n_nodes}
-#SBATCH --time=00:10:00
+#SBATCH --nodelist=f01[01-64]
+#SBATCH --time=00:30:00
 #SBATCH --export=NONE
 
+# Switch Help Table
+# SwitchName=fswibl01 Level=0 LinkSpeed=1 Nodes=f01[01-64]
+# SwitchName=fswibl02 Level=0 LinkSpeed=1 Nodes=f02[01-64]
+# SwitchName=fswibl03 Level=0 LinkSpeed=1 Nodes=f03[01-64]
+# SwitchName=fswibl04 Level=0 LinkSpeed=1 Nodes=f04[01-64]
+# SwitchName=fswibl05 Level=0 LinkSpeed=1 Nodes=f05[01-64]
+# SwitchName=fswibl06 Level=0 LinkSpeed=1 Nodes=f06[01-64]
+# SwitchName=fswibl07 Level=0 LinkSpeed=1 Nodes=f01[65-88],f02[65-88]
+# SwitchName=fswibl08 Level=0 LinkSpeed=1 Nodes=f03[65-88],f04[65-88],fritz[1-2]
+# SwitchName=fswibl09 Level=0 LinkSpeed=1 Nodes=f05[65-88],f06[65-88],fritz[3-4],fviz1
+# SwitchName=fswibl10 Level=0 LinkSpeed=1 Nodes=f07[01-64]
+# SwitchName=fswibl11 Level=0 LinkSpeed=1 Nodes=f08[01-64]
+# SwitchName=fswibl12 Level=0 LinkSpeed=1 Nodes=f09[01-64]
+# SwitchName=fswibl13 Level=0 LinkSpeed=1 Nodes=f10[01-64]
+
 unset SLURM_EXPORT_ENV
 
 module load intel intelmpi 
 
+# Enable tuned collectives
+export I_MPI_TUNING=on
+export I_MPI_TUNING_MODE=auto        # or 'collectives'
+
+# Options: 0=auto, 1=recursive doubling, 2=ring, 3=binomial tree, 4=scatter-allgather
+export I_MPI_COLL_ALLGATHER=2 
+export I_MPI_COLL_GATHER=2     
 
 OUTPUT_FILENAME="{data_dir}/{job_name}_$SLURM_JOB_ID.dat"
 
@@ -17,6 +40,5 @@ echo  "# CREATION_TIME : {time_stamp}" > $OUTPUT_FILENAME
 echo  "# N_NODES       : {n_nodes}" >> $OUTPUT_FILENAME
 echo  "# OFF_CACHE_FLAG  : {off_cache_flag}">> $OUTPUT_FILENAME
 
-srun --cpu-freq=2000000-2000000:performance -N {n_nodes} -n{n_procs} {bin} {job_name} -npmin {n_procs} {off_cache_flag}  >> $OUTPUT_FILENAME 
+srun --cpu-freq=2000000-2000000:performance -N {n_nodes} -n{n_procs} {bin} {job_name} -npmin {n_procs} {off_cache_flag} -mem 2 -time 60 >> $OUTPUT_FILENAME 
-
 

templates/multinode_algs.template

@@ -0,0 +1,35 @@
+#!/bin/bash -l
+#SBATCH --job-name={job_name}_{n_procs}
+#SBATCH --output={output_dir}{job_name}_{n_procs}.out
+#SBATCH --error={err_dir}{job_name}_{n_procs}.err
+#SBATCH --nodes={n_nodes}
+#SBATCH --nodelist=f01[01-64]
+#SBATCH --time=00:30:00
+#SBATCH --export=NONE
+# SwitchName=fswibl01 Level=0 LinkSpeed=1 Nodes=f01[01-64]
+# SwitchName=fswibl02 Level=0 LinkSpeed=1 Nodes=f02[01-64]
+# SwitchName=fswibl03 Level=0 LinkSpeed=1 Nodes=f03[01-64]
+# SwitchName=fswibl04 Level=0 LinkSpeed=1 Nodes=f04[01-64]
+# SwitchName=fswibl05 Level=0 LinkSpeed=1 Nodes=f05[01-64]
+# SwitchName=fswibl06 Level=0 LinkSpeed=1 Nodes=f06[01-64]
+# SwitchName=fswibl07 Level=0 LinkSpeed=1 Nodes=f01[65-88],f02[65-88]
+# SwitchName=fswibl08 Level=0 LinkSpeed=1 Nodes=f03[65-88],f04[65-88],fritz[1-2]
+# SwitchName=fswibl09 Level=0 LinkSpeed=1 Nodes=f05[65-88],f06[65-88],fritz[3-4],fviz1
+# SwitchName=fswibl10 Level=0 LinkSpeed=1 Nodes=f07[01-64]
+# SwitchName=fswibl11 Level=0 LinkSpeed=1 Nodes=f08[01-64]
+# SwitchName=fswibl12 Level=0 LinkSpeed=1 Nodes=f09[01-64]
+# SwitchName=fswibl13 Level=0 LinkSpeed=1 Nodes=f10[01-64]
+unset SLURM_EXPORT_ENV
+
+module load intel intelmpi 
+
+export I_MPI_ADJUST_{capital_jobname}={algnumber}
+
+OUTPUT_FILENAME="{data_dir}/{job_name}_$SLURM_JOB_ID.dat"
+
+echo  "# CREATION_TIME : {time_stamp}" > $OUTPUT_FILENAME
+echo  "# N_NODES       : {n_nodes}" >> $OUTPUT_FILENAME
+echo  "# OFF_CACHE_FLAG  : {off_cache_flag}">> $OUTPUT_FILENAME
+
+srun --cpu-freq=2000000-2000000:performance -N {n_nodes} -n{n_procs} {bin} {job_name} -npmin {n_procs} {off_cache_flag} -mem 2 -time 60 >> $OUTPUT_FILENAME 
+