# Julia at Scale

**URL:** https://discourse.julialang.org/c/domain/parallel/34.md?page=3

[Latest](https://discourse.julialang.org/latest.md) · [Categories](https://discourse.julialang.org/categories.md) · [Tags](https://discourse.julialang.org/tags.md)

**Page:** 4

---

## [Distributed Julia, ClusterMangers package, Slurm, and CUDA on a shared file system](https://discourse.julialang.org/t/distributed-julia-clustermangers-package-slurm-and-cuda-on-a-shared-file-system/90663)

<div class="topic-metadata">

**Author:** [@asaxton](https://discourse.julialang.org/u/asaxton)\
**Replies:** 4\
**Last updated:** [November 22, 2022, 8:27pm UTC](https://discourse.julialang.org/t/distributed-julia-clustermangers-package-slurm-and-cuda-on-a-shared-file-system/90663 "2022-11-22T20:27:36Z")

</div>

Hello, I’m running Julia 1.7.3 on a cluster using Slurm to launch batch jobs. The cluster has a shared filesystem. I’m using the Slurm cluster manager from the ClusterManagers package to call addproc(...). The node I’m …

---

## [Distributed.jl Processes not Shutting Down until Walltime on PBS](https://discourse.julialang.org/t/distributed-jl-processes-not-shutting-down-until-walltime-on-pbs/90388)

<div class="topic-metadata">

**Author:** [@vijeycreative](https://discourse.julialang.org/u/vijeycreative)\
**Replies:** 7\
**Last updated:** [November 19, 2022, 10:44am UTC](https://discourse.julialang.org/t/distributed-jl-processes-not-shutting-down-until-walltime-on-pbs/90388 "2022-11-19T10:44:09Z")

</div>

Hi All, This is my first time posting here, and I am posting after having read through a number of similar posts on Julia Discourse. My problem in a nutshell is that when I run a script on PBS, where the script does som…

---

## [MPI.API.MPI\_Reduce\_scatter ERRORS](https://discourse.julialang.org/t/mpi-api-mpi-reduce-scatter-errors/90093)

<div class="topic-metadata">

**Author:** [@Neo](https://discourse.julialang.org/u/Neo)\
**Replies:** 7\
**Last updated:** [November 15, 2022, 6:09am UTC](https://discourse.julialang.org/t/mpi-api-mpi-reduce-scatter-errors/90093 "2022-11-15T06:09:09Z")

</div>

From low-level API in MPI.jl, we know MPI.API.MPI\_Reduce\_scatter(sendbuf, recvbuf, recvcounts, datatype, op, comm) When I ran the code below with mpiexecjl -n 3 julia testReduce\_scatter.jl: # testReduce\_scatter.jl usi…

---

## [How to import low-level API in MPI.jl](https://discourse.julialang.org/t/how-to-import-low-level-api-in-mpi-jl/90086)

<div class="topic-metadata">

**Author:** [@Neo](https://discourse.julialang.org/u/Neo)\
**Replies:** 5\
**Last updated:** [November 11, 2022, 2:16am UTC](https://discourse.julialang.org/t/how-to-import-low-level-api-in-mpi-jl/90086 "2022-11-11T02:16:15Z")

</div>

I wanted to use one of the low-level APIs MPI.API.MPI\_Reduce\_scatter from MPI.jl. But I failed to import it using using MPI: API or using MPI MPI.API.MPI\_Reduce\_scatter(...) Does anyone know how to use it?

---

## [How to use Dagger in remote machines?](https://discourse.julialang.org/t/how-to-use-dagger-in-remote-machines/89793)

<div class="topic-metadata">

**Author:** [@Noel\_Araujo](https://discourse.julialang.org/u/Noel_Araujo)\
**Replies:** 3\
**Last updated:** [November 5, 2022, 4:07pm UTC](https://discourse.julialang.org/t/how-to-use-dagger-in-remote-machines/89793 "2022-11-05T16:07:34Z")

</div>

I began using Dagger.jl and I want to run some tests it with my local host machines, but those tests only runs in multithreading in my notebook. I know that my local machines work, because I can run them with Distribute…

---

## [Connection timed out problem on SSHManager of Distributed.jl](https://discourse.julialang.org/t/connection-timed-out-problem-on-sshmanager-of-distributed-jl/89129)

<div class="topic-metadata">

**Author:** [@hsianktin](https://discourse.julialang.org/u/hsianktin)\
**Replies:** 0\
**Last updated:** [October 23, 2022, 5:51am UTC](https://discourse.julialang.org/t/connection-timed-out-problem-on-sshmanager-of-distributed-jl/89129 "2022-10-23T05:51:57Z")

</div>

Hi everyone, I met a problem in adding another Windows machine as a worker using SSHManager. Background information: I am able to direct ssh into the remote machine, and the machine is under the same local network wit…

---

## [RemoteChannel based Collective Operations](https://discourse.julialang.org/t/remotechannel-based-collective-operations/88453)

<div class="topic-metadata">

**Author:** [@charlescolley](https://discourse.julialang.org/u/charlescolley)\
**Replies:** 0\
**Last updated:** [October 8, 2022, 4:57pm UTC](https://discourse.julialang.org/t/remotechannel-based-collective-operations/88453 "2022-10-08T16:57:49Z")

</div>

I have a small collection of tree-based collectives that I’m using for some of my projects. I’m new to distributed algorithms and wanted to get some thoughts on the code I’ve written and the profiling I’ve done in hopes …

---

## [\[ANN\] MPI.jl 0.20](https://discourse.julialang.org/t/ann-mpi-jl-0-20/88195)

<div class="topic-metadata">

**Author:** [@simonbyrne](https://discourse.julialang.org/u/simonbyrne)\
**Replies:** 0\
**Last updated:** [October 3, 2022, 10:37pm UTC](https://discourse.julialang.org/t/ann-mpi-jl-0-20/88195 "2022-10-03T22:37:33Z")

</div>

I’m happy to announce the latest release of MPI.jl, version 0.20. It is our hope that this will serve as a release candidate for future version 1.0. The biggest change by far in this release is the configuration procedu…

---

## [Load file on workers after they start](https://discourse.julialang.org/t/load-file-on-workers-after-they-start/87502)

<div class="topic-metadata">

**Author:** [@Amit.Rotem](https://discourse.julialang.org/u/Amit.Rotem)\
**Replies:** 0\
**Last updated:** [September 20, 2022, 5:41am UTC](https://discourse.julialang.org/t/load-file-on-workers-after-they-start/87502 "2022-09-20T05:41:03Z")

</div>

Is it possible to load a jl file on a worker as they start? The -L flag does not seems to work. using Distributed open("worker\_startup.jl", "w") do io print(io, """ using Distributed id = Distribute…

---

## [Difference between nprocs and nthreads](https://discourse.julialang.org/t/difference-between-nprocs-and-nthreads/87361)

<div class="topic-metadata">

**Author:** [@HenriDeh](https://discourse.julialang.org/u/HenriDeh)\
**Replies:** 2\
**Last updated:** [September 17, 2022, 5:35pm UTC](https://discourse.julialang.org/t/difference-between-nprocs-and-nthreads/87361 "2022-09-17T17:35:58Z")

</div>

Hello, I’m trying to get my head around distributed programming to execute a script on a cluster. My code works well on my workstation with one CPU. It internally uses @threads to parallelize some stuff within the main…

---

## [How to distribute arrays part of struct with MPI (MPI.jl)](https://discourse.julialang.org/t/how-to-distribute-arrays-part-of-struct-with-mpi-mpi-jl/87305)

<div class="topic-metadata">

**Author:** [@verseve](https://discourse.julialang.org/u/verseve)\
**Replies:** 0\
**Last updated:** [September 15, 2022, 8:30am UTC](https://discourse.julialang.org/t/how-to-distribute-arrays-part-of-struct-with-mpi-mpi-jl/87305 "2022-09-15T08:30:32Z")

</div>

I am wondering what the best approach is to distribute arrays that are part of a struct with MPI. Example of such a struct (partly): @get\_units @with\_kw struct SBM{T,N,M} Δt::T | "s" maxlayers::Int | "-" n:…

---

## [Help on terminology: multiprocessor vs. multicore](https://discourse.julialang.org/t/help-on-terminology-multiprocessor-vs-multicore/87066)

<div class="topic-metadata">

**Author:** [@HJW019](https://discourse.julialang.org/u/HJW019)\
**Replies:** 12\
**Last updated:** [September 15, 2022, 4:25am UTC](https://discourse.julialang.org/t/help-on-terminology-multiprocessor-vs-multicore/87066 "2022-09-15T04:25:46Z")

</div>

I am trying to understand the definitions of “multiprocessors” and “multicores” and it seems different sources have different definitions. I am mainly confused by whether the terms are defined by the numbers of physical …

---

## [Using local module with @everywhere](https://discourse.julialang.org/t/using-local-module-with-everywhere/87163)

<div class="topic-metadata">

**Author:** [@bojeryd91](https://discourse.julialang.org/u/bojeryd91)\
**Replies:** 2\
**Last updated:** [September 12, 2022, 9:38pm UTC](https://discourse.julialang.org/t/using-local-module-with-everywhere/87163 "2022-09-12T21:38:31Z")

</div>

Hi, my goal is to within one script define a module only once in my local scope, and then read it to all my workers. However, I cannot figure out what to do. This is my MWE: using Distributed module myModule export f f…

---

## [Easy way to send custom function to distributed workers?](https://discourse.julialang.org/t/easy-way-to-send-custom-function-to-distributed-workers/22118)

<div class="topic-metadata">

**Author:** [@nickeubank](https://discourse.julialang.org/u/nickeubank)\
**Replies:** 3\
**Last updated:** [September 12, 2022, 8:35pm UTC](https://discourse.julialang.org/t/easy-way-to-send-custom-function-to-distributed-workers/22118 "2022-09-12T20:35:07Z")

</div>

Is there a way to pass a function to workers? e.g. using Distributed addprocs(4) function my\_func(x) return x ^ 2 end \[send my function to everyone\] pmap(my\_func, 1:100) I recognize I can do something like: @ev…

---

## [How can I pass the pointer of MPI communicator to c function?](https://discourse.julialang.org/t/how-can-i-pass-the-pointer-of-mpi-communicator-to-c-function/87070)

<div class="topic-metadata">

**Author:** [@dora\_ho](https://discourse.julialang.org/u/dora_ho)\
**Replies:** 3\
**Last updated:** [September 11, 2022, 8:55pm UTC](https://discourse.julialang.org/t/how-can-i-pass-the-pointer-of-mpi-communicator-to-c-function/87070 "2022-09-11T20:55:51Z")

</div>

I have a c function that requires me to pass a comm pointer to the function I converted the function using Clang.jl showed below function Comm(comm\_handle) ccall((:Comm, librarypath), Cint, (Ptr{Cvoid}), comm\_hand…

---

## [Distributed.jl, DistributedArray.jl with InfiniBand cluster](https://discourse.julialang.org/t/distributed-jl-distributedarray-jl-with-infiniband-cluster/80049)

<div class="topic-metadata">

**Author:** [@ultrapoci](https://discourse.julialang.org/u/ultrapoci)\
**Replies:** 19\
**Last updated:** [September 7, 2022, 3:51pm UTC](https://discourse.julialang.org/t/distributed-jl-distributedarray-jl-with-infiniband-cluster/80049 "2022-09-07T15:51:54Z")

</div>

I’m developing an algorithm using Distributed.jl and DistributedArrays.jl, and I’m supposed to run this algorithm on a cluster supercomputer. I’ve recently been told that it uses Infiniband to connect internal nodes with…

---

## [Best way to parallelize](https://discourse.julialang.org/t/best-way-to-parallelize/85992)

<div class="topic-metadata">

**Author:** [@Dario-Rosa85](https://discourse.julialang.org/u/Dario-Rosa85)\
**Replies:** 12\
**Last updated:** [August 25, 2022, 2:01pm UTC](https://discourse.julialang.org/t/best-way-to-parallelize/85992 "2022-08-25T14:01:33Z")

</div>

Hi all, I have a large sparse matrix (dimension of order 10^5) which I have to multiply with a matrix of vectors many times. such a matrix of vectors is made out of, say, 500 vectors each of dimension equal to the dimen…

---

## [How to load ScaLAPACK library from Intel MKL (library inter-dependencies)?](https://discourse.julialang.org/t/how-to-load-scalapack-library-from-intel-mkl-library-inter-dependencies/81239)

<div class="topic-metadata">

**Author:** [@Sideboard](https://discourse.julialang.org/u/Sideboard)\
**Replies:** 4\
**Last updated:** [August 22, 2022, 8:23am UTC](https://discourse.julialang.org/t/how-to-load-scalapack-library-from-intel-mkl-library-inter-dependencies/81239 "2022-08-22T08:23:42Z")

</div>

Hi, I am trying to use ScaLAPACK with @ccall to solve a linear system with QR decomposition. On my laptop it works fine with OpenMPI and directly searching for the default ScaLAPACK library: const libscalapack = Base.L…

---

## [Poor scaling results with \`ImplicitGlobalGrid.jl\`](https://discourse.julialang.org/t/poor-scaling-results-with-implicitglobalgrid-jl/65170)

<div class="topic-metadata">

**Author:** [@francispoulin](https://discourse.julialang.org/u/francispoulin)\
**Replies:** 28\
**Last updated:** [August 3, 2022, 12:34pm UTC](https://discourse.julialang.org/t/poor-scaling-results-with-implicitglobalgrid-jl/65170 "2022-08-03T12:34:49Z")

</div>

I have tried running one of the examples of ImplicitGlobalGrid.jl on 1, 2 and 4 gpus to reproduce some scaling results. I started wtih diffusion3D\_multigpu\_CuArray\_novis.jl with a coarser grid to see how it does. Below y…

---

## [Graph computing benchmarks: comparing the scalability of Dask, Dagger.jl, Tensorflow and Julius](https://discourse.julialang.org/t/graph-computing-benchmarks-comparing-the-scalability-of-dask-dagger-jl-tensorflow-and-julius/80745)

<div class="topic-metadata">

**Author:** [@Yadong\_Li](https://discourse.julialang.org/u/Yadong_Li)\
**Replies:** 6\
**Last updated:** [July 22, 2022, 3:29pm UTC](https://discourse.julialang.org/t/graph-computing-benchmarks-comparing-the-scalability-of-dask-dagger-jl-tensorflow-and-julius/80745 "2022-07-22T15:29:55Z")

</div>

Graph computing solutions like Dask/Dagger.jl are gaining popularity among developers, mainly because of their easy built-in distribution capabilities. If you are considering one of these graph computing solutions for yo…

---

## [Why is the threaded version of this so much slower than the serial and distributed versions?](https://discourse.julialang.org/t/why-is-the-threaded-version-of-this-so-much-slower-than-the-serial-and-distributed-versions/83280)

<div class="topic-metadata">

**Author:** [@ElOceanografo](https://discourse.julialang.org/u/ElOceanografo)\
**Replies:** 2\
**Last updated:** [June 23, 2022, 10:23pm UTC](https://discourse.julialang.org/t/why-is-the-threaded-version-of-this-so-much-slower-than-the-serial-and-distributed-versions/83280 "2022-06-23T22:23:39Z")

</div>

Minimal example of my problem: using ThreadsX using Statistics data = \[randn(1000) for i in 1:100, j in 1:200\] f(x) = mean(x) # warmup f(data\[1, 2\]) t\_serial = @elapsed map(f, data) t\_thread = @elapsed ThreadsX.map(…

---

## [\[ANN\] Training course "Julia for HPC" in Stuttgart, Germany on Sep 20-23, 2022](https://discourse.julialang.org/t/ann-training-course-julia-for-hpc-in-stuttgart-germany-on-sep-20-23-2022/83111)

<div class="topic-metadata">

**Author:** [@sloede](https://discourse.julialang.org/u/sloede)\
**Replies:** 3\
**Last updated:** [June 22, 2022, 7:20am UTC](https://discourse.julialang.org/t/ann-training-course-julia-for-hpc-in-stuttgart-germany-on-sep-20-23-2022/83111 "2022-06-22T07:20:07Z")

</div>

Hey everyone! At HLRS we are offering a new “Julia for High-Performance Computing” course, which will take place on-site at the High-Performance Computing Center Stuttgart (HLRS) in Stuttgart, Germany from September 20t…

---

## [HPC systems with Julia support](https://discourse.julialang.org/t/hpc-systems-with-julia-support/82951)

<div class="topic-metadata">

**Author:** [@sloede](https://discourse.julialang.org/u/sloede)\
**Replies:** 4\
**Last updated:** [June 19, 2022, 10:30am UTC](https://discourse.julialang.org/t/hpc-systems-with-julia-support/82951 "2022-06-19T10:30:23Z")

</div>

Hey everyone! I am trying to collect information about HPC systems (from university compute clusters to supercomputers) where Julia is available as a pre-installed module and, optimally, officially supported. This is par…

---

## [Parallel reductions](https://discourse.julialang.org/t/parallel-reductions/30180)

<div class="topic-metadata">

**Author:** [@cscherrer](https://discourse.julialang.org/u/cscherrer)\
**Replies:** 23\
**Last updated:** [June 19, 2022, 2:43am UTC](https://discourse.julialang.org/t/parallel-reductions/30180 "2022-06-19T02:43:50Z")

</div>

“Parallel mapreduce” is a really common need, and it’s surprisingly hard to find a Julia implementation. Say we have julia\> x = randn(10000000); julia\> f = log∘sqrt∘abs #56 (generic function with 1 method) The goal i…

---

## [Adding Threads](https://discourse.julialang.org/t/adding-threads/36011)

<div class="topic-metadata">

**Author:** [@marouane1994](https://discourse.julialang.org/u/marouane1994)\
**Replies:** 4\
**Last updated:** [June 10, 2022, 2:01pm UTC](https://discourse.julialang.org/t/adding-threads/36011 "2022-06-10T14:01:24Z")

</div>

i’m newbie parallel computing and i’m trying to practice it a little bit and when i want to add threads: set JULIA\_NUM\_THREADS =4 i get this Error and i don’t get if i should add something or what: ERROR: syntax: extra t…

---

## [Save each simulation result on master node during parameter scanning](https://discourse.julialang.org/t/save-each-simulation-result-on-master-node-during-parameter-scanning/81953)

<div class="topic-metadata">

**Author:** [@jtoda](https://discourse.julialang.org/u/jtoda)\
**Replies:** 0\
**Last updated:** [May 31, 2022, 2:31am UTC](https://discourse.julialang.org/t/save-each-simulation-result-on-master-node-during-parameter-scanning/81953 "2022-05-31T02:31:53Z")

</div>

Hello ! I want to write code that stores data for a huge amount of simulation. I want to save data for each experiment, and the data storage location is the master node. Below is an example code, do you have any bette…

---

## [Code that works fine distributed across processes on one node using slurm seems to fail when trying to generate workers across many](https://discourse.julialang.org/t/code-that-works-fine-distributed-across-processes-on-one-node-using-slurm-seems-to-fail-when-trying-to-generate-workers-across-many/39359)

<div class="topic-metadata">

**Author:** [@HarrisonWilde](https://discourse.julialang.org/u/HarrisonWilde)\
**Replies:** 2\
**Last updated:** [May 19, 2022, 4:15pm UTC](https://discourse.julialang.org/t/code-that-works-fine-distributed-across-processes-on-one-node-using-slurm-seems-to-fail-when-trying-to-generate-workers-across-many/39359 "2022-05-19T16:15:51Z")

</div>

Hi, I have some code that works fine when I run it on a single node of our slurm cluster using sbatch, but when I try to run it on multiple nodes I get errors similar to the following: srun: error: cnode22: tasks 7,12,1…

---

## [Does anyone run Singularity images across many Nodes?](https://discourse.julialang.org/t/does-anyone-run-singularity-images-across-many-nodes/60470)

<div class="topic-metadata">

**Author:** [@Noel\_Araujo](https://discourse.julialang.org/u/Noel_Araujo)\
**Replies:** 2\
**Last updated:** [May 4, 2022, 10:07pm UTC](https://discourse.julialang.org/t/does-anyone-run-singularity-images-across-many-nodes/60470 "2022-05-04T22:07:04Z")

</div>

As I point out in here, I wish to connect ClusterManagers.jl using Singularity images, but as @kescobo says, the problem seems to be Singularity or my Cluster configuration. Does anyone already have experience running a…

---

## [Loading packages to the freshly added processes](https://discourse.julialang.org/t/loading-packages-to-the-freshly-added-processes/80434)

<div class="topic-metadata">

**Author:** [@Gregstrq](https://discourse.julialang.org/u/Gregstrq)\
**Replies:** 3\
**Last updated:** [May 3, 2022, 5:49pm UTC](https://discourse.julialang.org/t/loading-packages-to-the-freshly-added-processes/80434 "2022-05-03T17:49:07Z")

</div>

I am trying to write the function that launches multithreaded processes and then pins the threads to the core. The stripped down version looks like using Distributed, ThreadPinning function add\_pinned\_workers(nw::Int, …

---

## [Parallel HDF5 crashing on build](https://discourse.julialang.org/t/parallel-hdf5-crashing-on-build/80135)

<div class="topic-metadata">

**Author:** [@Chiil](https://discourse.julialang.org/u/Chiil)\
**Replies:** 1\
**Last updated:** [May 2, 2022, 12:07pm UTC](https://discourse.julialang.org/t/parallel-hdf5-crashing-on-build/80135 "2022-05-02T12:07:31Z")

</div>

I am using a MPI-enabled HDF5 and suddenly calling up from the Julia package manager fails with ERROR: Error building \`HDF5\`: signal (11): Segmentation fault: 11 in expression starting at /Users/chiel/.julia/packages/…

[Previous page](https://discourse.julialang.org/c/domain/parallel/34.md?page=2)

[Next page](https://discourse.julialang.org/c/domain/parallel/34.md?page=4)