Differenze

Queste sono le differenze tra la revisione selezionata e la versione attuale della pagina.

--- calcoloscientifico:userguide [15/04/2025 22:54] – fabio.spataro
+++ calcoloscientifico:userguide [21/05/2025 17:34] (versione attuale) – [Table] fabio.spataro
@@ Linea 143: / Linea 143: @@
 | wn48       | NVIDIA Corporation Device 2231 [RTX A5000 24GB] (rev a1)    |
 | wn49-wn50  | NVIDIA Corporation Device 26b9 [L40S 48GB] (rev a1)         |
+| wn51       | NVIDIA Corporation Device 2321 [H100 NVL 96GB] (rev a1)     |
 **Peak Performance**
@@ Linea 269: / Linea 270: @@
 | cpu_guest       | cpu_guest       | wn[20-26]              |   440 core |   1-00:00:00 |           24 |         200 |          4 | Guest     |
 | gpu             | gpu             | wn[41-42,44-45]        |     24 gpu |   1-00:00:00 |            6 |        2000 |          4 | Public    |
-| gpu_vbd         | gpu_vbd         | wn[46,49]              |     12 gpu |  14-00:00:00 |            6 |        2000 |          2 | Private   |
+| gpu_vbd         | gpu_vbd         | wn[46,49,51]           |     14 gpu |  14-00:00:00 |            6 |        2000 |          2 | Private   |
 | gpu_hylab       | gpu_hylab       | wn44                   |      4 gpu |   3-00:00:00 |            4 |        2000 |          1 | Private   |
+| gpu_fisstat     | gpu_fisstat     | wn50                   |      4 gpu |   7-00:00:00 |            4 |        2000 |          1 | Private   |
 | gpu_guest       | gpu_guest       | wn[44,46]              |      8 gpu |   1-00:00:00 |            6 |         200 |          2 | Guest     |
-| knl             | knl             | wn[51-54]              |  1088 core |  14-00:00:00 |           16 |        2000 |          4 | Public    |
 | vrt             | vrt             | wn[61-64]              |     1 core |  10-00:00:00 |           24 |        2000 |          1 | Public    |
 | mngt            |                 | all                    |            | UNLIMITED    |              |             |            | Admin     |
@@ Linea 328: / Linea 329: @@
 hpc-sinfo-cpu --partition=cpu
 hpc-sinfo-cpu --partition=cpu_guest
-hpc-sinfo-cpu --partition=knl
 hpc-sinfo-gpu --partition=gpu
 hpc-sinfo-gpu --partition=gpu_guest
@@ Linea 340: / Linea 340: @@
 hpc-squeue --partition=cpu
 hpc-squeue --partition=cpu_guest
-hpc-squeue --partition=knl
 hpc-squeue --partition=gpu
 hpc-squeue --partition=gpu_guest
@@ Linea 609: / Linea 608: @@
 <code>
 srun --nodes=1 --nodelist=wn34 --pty bash
-</code>
-Reserve 1 chunk of 272 CPUs type KNL (a whole KNL node):
-<code bash>
-srun --nodes=1 --ntasks-per-node=272 --partition=knl --qos=knl --mem=8G --pty bash
 </code>
@@ Linea 1136: / Linea 1130: @@
 sbatch matrix-vector-product.sh
 </code>
-=====  KNL jobs =====
-The compiler to be used is Intel.
-The selection of the KNL cluster is done by specifying ''-p knl''.
-Each knl node includes 68  physical core and each physical core includes 4 virtual cores with hyperthreading technology. The maximum number of cores (ncpus) selectable per node is 272.
-MPI + OpenMP example on KNL:
-   cp /hpc/share/samples/mpi+omp/mpiomp_hello.c .
-<code c mpiomp_hello.c>
-#include <stdio.h>
-#include <mpi.h>
-#include <omp.h>
-int main(int argc, char *argv[]) {
-  int numprocs, rank, namelen;
-  char processor_name[MPI_MAX_PROCESSOR_NAME];
-  int iam = 0, np = 1;
-  MPI_Init(&argc, &argv);
-  MPI_Comm_size(MPI_COMM_WORLD, &numprocs);
-  MPI_Comm_rank(MPI_COMM_WORLD, &rank);
-  MPI_Get_processor_name(processor_name, &namelen);
-#pragma omp parallel default(shared) private(iam, np)
-  {
-    np = omp_get_num_threads();
-    iam = omp_get_thread_num();
-    printf("Hello from thread %3d out of %3d from process %3d out of %3d on %s\n",
-           iam + 1, np, rank + 1, numprocs, processor_name);
-  }
-  MPI_Finalize();
-}
-</code>
-Sumbission script ''mpiomp_hello_knl.bash''
-<code bash mpiomp_hello_knl.bash>
-#!/bin/sh
-#< 2  nodes. Executes 1 MPI process per node and 68 threads per process
-#SBATCH --partition=knl
-#SBATCH --qos=knl
-#SBATCH --nodes=2
-#SBATCH --ntasks=2
-#SBATCH --cpus-per-task=68      #< Number of threads OpenMP for each process MPI
-#SBATCH --time=0-00:30:00
-#SBATCH --mem=4G
-## Uncomment the following line if you want to use an account other than your default account
-##SBATCH --account=<account>
-export OMP_NUM_THREADS=$SLURM_CPUS_PER_TASK
-module load intel intelmpi
-mpicc -qopenmp mpiomp_hello.c -o mpiomp_hello_knl
-mpirun mpiomp_hello_knl
-</code>
-===Submission===
-  sbatch mpiomp_hello_knl.bash
 ===== GPU jobs =====