While these tutorials are typically meant to be visualized in serial, viskex also runs in parallel. To exemplify this, we will show a parallel case using ipyparallel.
First of all, we start a ipyparallel MPI cluster with 2 processes.
import ipyparallel as ipp
cluster = ipp.Cluster(engines="MPI", profile="mpi", n=2)
cluster.start_and_connect_sync()
Starting 2 engines with <class 'ipyparallel.cluster.launcher.MPIEngineSetLauncher'>
<ipyparallel.client.client.Client at 0x7fcd005dd3a0>
The jupyter magic %%px will run the following cells on the ipyparallel cluster.
%%px
import firedrake # noqa: E402
import mpi4py.MPI # noqa: E402
%%px
import viskex # noqa: E402
To confirm that we are running in parallel, we can print the rank of the current process and the total number of processes.
%%px
comm_world = mpi4py.MPI.COMM_WORLD
print(f"rank = {comm_world.rank}, size = {comm_world.size}")
[stdout:1] rank = 1, size = 2
[stdout:0] rank = 0, size = 2
Generate a mesh on MPI_COMM_WORLD of the unit square by dividing each edge of the square in 6 segments.
%%px
square_world = firedrake.UnitSquareMesh(
6, 6, comm=comm_world, distribution_parameters={"partitioner_type": "simple"})
Plot the mesh defined on MPI_COMM_WORLD. Each rank will plot its local cells, and their neighbors (called halos in firedrake).
%%px
viskex.firedrake.plot_mesh(square_world)
[stderr:0] 2026-03-23 05:07:23.668 ( 1.110s) [ 7FEB37583080]vtkXOpenGLRenderWindow.:1460 WARN| bad X server connection. DISPLAY=
[stderr:1] 2026-03-23 05:07:23.668 ( 1.107s) [ 7FD2F5145080]vtkXOpenGLRenderWindow.:1460 WARN| bad X server connection. DISPLAY=
[output:0]
[output:1]
For comparison, we can generate a similar mesh on MPI_COMM_SELF. Each rank will have its own copy of the whole mesh.
%%px
comm_self = mpi4py.MPI.COMM_SELF
square_self = firedrake.UnitSquareMesh(
6, 6, comm=comm_self, distribution_parameters={"partitioner_type": "simple"})
Plot the mesh defined on MPI_COMM_SELF. Each rank will produce a plot which is visually the same, since each rank has its own copy of the same mesh.
%%px
viskex.firedrake.plot_mesh(square_self)
[output:0]
[output:1]
We finally stop the ipyparallel cluster. Note that when running with Run -> Run all cells all interactive plots will be closed when the kernel executes this cell.
cluster.stop_cluster_sync()
Stopping controller
Controller stopped: {'exit_code': 0, 'pid': 2763, 'identifier': 'ipcontroller-1774242434-axjk-2744'}
Stopping engine(s): 1774242435
Output for ipengine-1774242434-axjk-1774242435-2744: 2026-03-23 05:07:16.212 [IPEngine] Loading connection info from $IPP_CONNECTION_INFO 2026-03-23 05:07:16.212 [IPEngine] WARNING | Not using CurveZMQ security 2026-03-23 05:07:16.228 [IPEngine] Loading connection info from $IPP_CONNECTION_INFO 2026-03-23 05:07:16.228 [IPEngine] WARNING | Not using CurveZMQ security 2026-03-23 05:07:16.461 [IPEngine.0] Registering with controller at tcp://127.0.0.1:55979 2026-03-23 05:07:16.461 [IPEngine.1] Registering with controller at tcp://127.0.0.1:55979 2026-03-23 05:07:16.462 [IPEngine.1] Requesting id: 1 2026-03-23 05:07:16.462 [IPEngine.0] Requesting id: 0 2026-03-23 05:07:16.464 [IPEngine.1.1] Shell_addrs: ['tcp://127.0.0.1:42855', 'tcp://127.0.0.1:39065', 'tcp://127.0.0.1:53851'] 2026-03-23 05:07:16.464 [IPEngine.1.1] Connecting shell to tcp://127.0.0.1:42855 2026-03-23 05:07:16.465 [IPEngine.1.1] Connecting shell to tcp://127.0.0.1:39065 2026-03-23 05:07:16.465 [IPEngine.1.1] Connecting shell to tcp://127.0.0.1:53851 2026-03-23 05:07:16.465 [IPEngine.1.1] Starting nanny 2026-03-23 05:07:16.465 [IPEngine.0.0] Shell_addrs: ['tcp://127.0.0.1:42855', 'tcp://127.0.0.1:39065', 'tcp://127.0.0.1:57337'] 2026-03-23 05:07:16.466 [IPEngine.0.0] Connecting shell to tcp://127.0.0.1:42855 2026-03-23 05:07:16.466 [IPEngine.0.0] Connecting shell to tcp://127.0.0.1:39065 2026-03-23 05:07:16.466 [IPEngine.0.0] Connecting shell to tcp://127.0.0.1:57337 2026-03-23 05:07:16.466 [IPEngine.0.0] Starting nanny 2026-03-23 05:07:16.900 [KernelNanny.1] Starting kernel nanny for engine 1, pid=2799, nanny pid=2809 2026-03-23 05:07:16.900 [KernelNanny.1] Nanny watching parent pid 2799. 2026-03-23 05:07:16.905 [KernelNanny.0] Starting kernel nanny for engine 0, pid=2798, nanny pid=2810 2026-03-23 05:07:16.905 [KernelNanny.0] Nanny watching parent pid 2798. 2026-03-23 05:07:16.968 [IPEngine.1.1] Loading IPython extension: storemagic 2026-03-23 05:07:16.969 [IPEngine.1.1] Running code in user namespace: from mpi4py import MPI mpi_rank = MPI.COMM_WORLD.Get_rank() mpi_size = MPI.COMM_WORLD.Get_size() 2026-03-23 05:07:16.970 [IPEngine.1.1] WARNING | debugpy_stream undefined, debugging will not be enabled 2026-03-23 05:07:16.972 [IPEngine.1.1] Starting to monitor the heartbeat signal from the hub every 3500 ms. 2026-03-23 05:07:16.972 [IPEngine.1.1] Completed registration with id 1 2026-03-23 05:07:16.974 [IPEngine.0.0] Loading IPython extension: storemagic 2026-03-23 05:07:16.975 [IPEngine.0.0] Running code in user namespace: from mpi4py import MPI mpi_rank = MPI.COMM_WORLD.Get_rank() mpi_size = MPI.COMM_WORLD.Get_size() 2026-03-23 05:07:16.976 [IPEngine.0.0] WARNING | debugpy_stream undefined, debugging will not be enabled 2026-03-23 05:07:16.978 [IPEngine.0.0] Starting to monitor the heartbeat signal from the hub every 3500 ms. 2026-03-23 05:07:16.978 [IPEngine.0.0] Completed registration with id 0 2026-03-23 05:07:21.463 [IPEngine.0.0] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_1 2026-03-23 05:07:21.463 [IPEngine.1.1] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_2 2026-03-23 05:07:22.891 [IPEngine.0.0] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_3 2026-03-23 05:07:22.892 [IPEngine.1.1] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_4 2026-03-23 05:07:23.587 [IPEngine.0.0] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_5 2026-03-23 05:07:23.587 [IPEngine.1.1] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_6 2026-03-23 05:07:23.598 [IPEngine.0.0] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_7 2026-03-23 05:07:23.598 [IPEngine.1.1] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_8 2026-03-23 05:07:23.659 [IPEngine.0.0] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_9 2026-03-23 05:07:23.659 [IPEngine.1.1] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_10 2026-03-23 05:07:23.668 ( 1.110s) [ 7FEB37583080]vtkXOpenGLRenderWindow.:1460 WARN| bad X server connection. DISPLAY= 2026-03-23 05:07:23.668 ( 1.107s) [ 7FD2F5145080]vtkXOpenGLRenderWindow.:1460 WARN| bad X server connection. DISPLAY= 2026-03-23 05:07:24.467 [IPEngine.0.0] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_11 2026-03-23 05:07:24.468 [IPEngine.1.1] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_12 2026-03-23 05:07:24.484 [IPEngine.0.0] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_13 2026-03-23 05:07:24.484 [IPEngine.1.1] Handling execute_request: b908d4da-20aa541e8862440ac6c625eb_2744_14
engine set stopped 1774242435: {'exit_code': 1, 'pid': 2791, 'identifier': 'ipengine-1774242434-axjk-1774242435-2744'}