# Julia at Scale

**URL:** https://discourse.julialang.org/c/domain/parallel/34.md?page=2

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

**Page:** 3

---

## [Odd thread calling behavior on HPC](https://discourse.julialang.org/t/odd-thread-calling-behavior-on-hpc/104789)

<div class="topic-metadata">

**Author:** [@pat-alt](https://discourse.julialang.org/u/pat-alt)\
**Replies:** 2\
**Last updated:** [October 11, 2023, 12:23pm UTC](https://discourse.julialang.org/t/odd-thread-calling-behavior-on-hpc/104789 "2023-10-11T12:23:24Z")

</div>

I’ve been running Julia on my university’s HPC using a self-installed version of Julia (through juliaup) and have noticed odd thread calling behavior. When running the following Julia script a = zeros(10) Threads.@thre…

---

## [Distributing a job across a cluster with SlurmClusterManager.jl](https://discourse.julialang.org/t/distributing-a-job-across-a-cluster-with-slurmclustermanager-jl/104735)

<div class="topic-metadata">

**Author:** [@gideonsimpson](https://discourse.julialang.org/u/gideonsimpson)\
**Replies:** 2\
**Last updated:** [October 10, 2023, 7:41pm UTC](https://discourse.julialang.org/t/distributing-a-job-across-a-cluster-with-slurmclustermanager-jl/104735 "2023-10-10T19:41:22Z")

</div>

I asked about this on the slack channel, but didn’t get a response. I’ve been using SlurmClusterManager and been pretty satisfied with it, but there’s a job I want to accomplish with it that I’m not sure how to do. The…

---

## [Unexpected OOM errrors in julia 1.9.0 and 1.9.1 with Distributed](https://discourse.julialang.org/t/unexpected-oom-errrors-in-julia-1-9-0-and-1-9-1-with-distributed/100103)

<div class="topic-metadata">

**Author:** [@vfonov](https://discourse.julialang.org/u/vfonov)\
**Replies:** 6\
**Last updated:** [September 28, 2023, 2:38am UTC](https://discourse.julialang.org/t/unexpected-oom-errrors-in-julia-1-9-0-and-1-9-1-with-distributed/100103 "2023-09-28T02:38:18Z")

</div>

I have a relatively simple script that used to run fine on 1.8.X version: it executes lots of relatively small jobs via pmap. Executing on julia 1.8.5 with -p 32 , creates 32 processes , each using approximately 2Gb an…

---

## [Speed issues with MPI.jl on slurm cluster](https://discourse.julialang.org/t/speed-issues-with-mpi-jl-on-slurm-cluster/103068)

<div class="topic-metadata">

**Author:** [@gerhardu](https://discourse.julialang.org/u/gerhardu)\
**Replies:** 4\
**Last updated:** [August 25, 2023, 9:55am UTC](https://discourse.julialang.org/t/speed-issues-with-mpi-jl-on-slurm-cluster/103068 "2023-08-25T09:55:49Z")

</div>

Hello, I wrote MPI.jl parallel code (my first one) to be used on a Slurm cluster. I ran the code on my notebook and on the cluster using one node. It turns out that the code executes 6 times slower on the cluster and …

---

## [Some Dagger.jl beginner questions](https://discourse.julialang.org/t/some-dagger-jl-beginner-questions/102704)

<div class="topic-metadata">

**Author:** [@fabiangans](https://discourse.julialang.org/u/fabiangans)\
**Replies:** 6\
**Last updated:** [August 23, 2023, 9:22pm UTC](https://discourse.julialang.org/t/some-dagger-jl-beginner-questions/102704 "2023-08-23T21:22:22Z")

</div>

Hi all, I recently started working with Dagger.jl and have a few questions about controlling the scheduler and merging shards. In my workflow I am doing a mapreduce-like operation over hierarchical data. An abstract ver…

---

## [Run Pkg.test with several processors](https://discourse.julialang.org/t/run-pkg-test-with-several-processors/6917)

<div class="topic-metadata">

**Author:** [@Wikunia](https://discourse.julialang.org/u/Wikunia)\
**Replies:** 16\
**Last updated:** [August 21, 2023, 9:48am UTC](https://discourse.julialang.org/t/run-pkg-test-with-several-processors/6917 "2023-08-21T09:48:50Z")

</div>

I have tests that use several processors. If I run julia -p 4 and then Pkg.test I only get 1 as a worker.

---

## [Distributed Array Optimization](https://discourse.julialang.org/t/distributed-array-optimization/101807)

<div class="topic-metadata">

**Author:** [@Brian\_Makin](https://discourse.julialang.org/u/Brian_Makin)\
**Replies:** 0\
**Last updated:** [July 19, 2023, 9:27pm UTC](https://discourse.julialang.org/t/distributed-array-optimization/101807 "2023-07-19T21:27:39Z")

</div>

I’ve got a distributed array split across a bunch of workers. Each worker grabs portions of the distributed array (which creates subarrays) they processes their local part while accessing those subarrays. Two question…

---

## [Changing default interface for Julia worker](https://discourse.julialang.org/t/changing-default-interface-for-julia-worker/101746)

<div class="topic-metadata">

**Author:** [@JasperW99](https://discourse.julialang.org/u/JasperW99)\
**Replies:** 1\
**Last updated:** [July 18, 2023, 9:20pm UTC](https://discourse.julialang.org/t/changing-default-interface-for-julia-worker/101746 "2023-07-18T21:20:09Z")

</div>

I am running a distributed workload using Distributed.jl on a cluster that has set up IP over Infiniband (IPoIB). However, when the workers are created, they automatically return the IP address of the slow ethernet inter…

---

## [\[ANN\] Training course “Julia for HPC” in Stuttgart, Germany on Oct 24-27, 2023](https://discourse.julialang.org/t/ann-training-course-julia-for-hpc-in-stuttgart-germany-on-oct-24-27-2023/101251)

<div class="topic-metadata">

**Author:** [@sloede](https://discourse.julialang.org/u/sloede)\
**Replies:** 1\
**Last updated:** [July 11, 2023, 8:22am UTC](https://discourse.julialang.org/t/ann-training-course-julia-for-hpc-in-stuttgart-germany-on-oct-24-27-2023/101251 "2023-07-11T08:22:15Z")

</div>

After the great positive feedback we received at last year’s Julia for HPC training course at HLRS, there will be another opportunity to learn more about using Julia for high-performance computing this fall! At HLRS, we…

---

## [Hyperthreading in HPC](https://discourse.julialang.org/t/hyperthreading-in-hpc/101316)

<div class="topic-metadata">

**Author:** [@fgerick](https://discourse.julialang.org/u/fgerick)\
**Replies:** 16\
**Last updated:** [July 8, 2023, 7:04am UTC](https://discourse.julialang.org/t/hyperthreading-in-hpc/101316 "2023-07-08T07:04:18Z")

</div>

Hi everyone, I have a question regarding the use of hyperthreading/SMT on HPC systems. HPC hyperthreading on off - Google Search reveals some vague hints that in the past it has been accepted that enabling hyperthreadin…

---

## [Distributed.jl: Group myid() by computing nodes?](https://discourse.julialang.org/t/distributed-jl-group-myid-by-computing-nodes/100619)

<div class="topic-metadata">

**Author:** [@lmtzx9h4qqnt](https://discourse.julialang.org/u/lmtzx9h4qqnt)\
**Replies:** 3\
**Last updated:** [June 23, 2023, 9:30am UTC](https://discourse.julialang.org/t/distributed-jl-group-myid-by-computing-nodes/100619 "2023-06-23T09:30:14Z")

</div>

Using Distributed.jl, is there no built-in way to find out which workers are on the same shared-memory node? This post: shows a manual implementation based on gethostname(), but I don’t know how robust that is? If ther…

---

## [Garbage collection not aggressive enough on Slurm Cluster](https://discourse.julialang.org/t/garbage-collection-not-aggressive-enough-on-slurm-cluster/61649)

<div class="topic-metadata">

**Author:** [@mkschleg](https://discourse.julialang.org/u/mkschleg)\
**Replies:** 17\
**Last updated:** [August 4, 2021, 2:11pm UTC](https://discourse.julialang.org/t/garbage-collection-not-aggressive-enough-on-slurm-cluster/61649 "2021-08-04T14:11:50Z")

</div>

I’ve been using slurm for awhile w/ julia. I’ve only recently been running into issues where I’m running out of memory for long running processes (like 30hrs). The jobs are highly regular, effectively calling the same fu…

---

## [Rust in HPC study](https://discourse.julialang.org/t/rust-in-hpc-study/99834)

<div class="topic-metadata">

**Author:** [@johnh](https://discourse.julialang.org/u/johnh)\
**Replies:** 0\
**Last updated:** [June 4, 2023, 7:24am UTC](https://discourse.julialang.org/t/rust-in-hpc-study/99834 "2023-06-04T07:24:30Z")

</div>

OK, specific to Rust. I will happily move to another category. however this may be interesting to the audience here. Quote: The three languages used are Rust, C, and Fortran and the programme is run for serial and paral…

---

## [How to run MPI jobs on a cluster](https://discourse.julialang.org/t/how-to-run-mpi-jobs-on-a-cluster/97191)

<div class="topic-metadata">

**Author:** [@Neo](https://discourse.julialang.org/u/Neo)\
**Replies:** 1\
**Last updated:** [June 2, 2023, 6:56pm UTC](https://discourse.julialang.org/t/how-to-run-mpi-jobs-on-a-cluster/97191 "2023-06-02T18:56:00Z")

</div>

I want to run Julia MPI jobs on a cluster but not sure if I did it in the correct way. There is a large cluster with hundreds of nodes where each node has 128 cores. The resource management system on the cluster is SLUR…

---

## [Higher order derivatives with ParallelStencil](https://discourse.julialang.org/t/higher-order-derivatives-with-parallelstencil/99647)

<div class="topic-metadata">

**Author:** [@Teerthal](https://discourse.julialang.org/u/Teerthal)\
**Replies:** 2\
**Last updated:** [May 31, 2023, 6:24pm UTC](https://discourse.julialang.org/t/higher-order-derivatives-with-parallelstencil/99647 "2023-05-31T18:24:01Z")

</div>

I am currently implementing a numerical simulation code on multiple GPUs using ParallelStencil+ImplicitGlobalGrid. I wanted to use higher order numerical differentiation schemes which I could easily do by extending the …

---

## [Distributed for loop with timeout](https://discourse.julialang.org/t/distributed-for-loop-with-timeout/98854)

<div class="topic-metadata">

**Author:** [@psterzinger](https://discourse.julialang.org/u/psterzinger)\
**Replies:** 1\
**Last updated:** [May 15, 2023, 5:16pm UTC](https://discourse.julialang.org/t/distributed-for-loop-with-timeout/98854 "2023-05-15T17:16:05Z")

</div>

I’m trying to run a function, which consists of a nested for loop, in which each iteration writes to a SharedArray, and where the runtime between iterations varies unpredictably, with a timeout such that upon termination…

---

## [MPI collectives (e.g., MPI.Allgatherv!) scaling issues](https://discourse.julialang.org/t/mpi-collectives-e-g-mpi-allgatherv-scaling-issues/97129)

<div class="topic-metadata">

**Author:** [@Neo](https://discourse.julialang.org/u/Neo)\
**Replies:** 1\
**Last updated:** [May 15, 2023, 2:33am UTC](https://discourse.julialang.org/t/mpi-collectives-e-g-mpi-allgatherv-scaling-issues/97129 "2023-05-15T02:33:10Z")

</div>

I found that MPI.Allgatherv! doesn’t scale well when testing it on a cluster. The plot above summarizes the execution time of MPI.Allgatherv! against the number of processes. The scaling shows in the plot doesn’t mat…

---

## [Distributed computing for functions in scripts inside a local module?](https://discourse.julialang.org/t/distributed-computing-for-functions-in-scripts-inside-a-local-module/96729)

<div class="topic-metadata">

**Author:** [@math\_opt](https://discourse.julialang.org/u/math_opt)\
**Replies:** 7\
**Last updated:** [May 5, 2023, 7:33pm UTC](https://discourse.julialang.org/t/distributed-computing-for-functions-in-scripts-inside-a-local-module/96729 "2023-05-05T19:33:23Z")

</div>

I have a local module, say MyProject, which has a folder src with two scripts script1.jl and script2.jl. module MyProject using Package1 using Package2 using Package3 include("script1.jl") include("script2.jl") expor…

---

## [Extend distributed pmap to mutliple nodes using SLURM](https://discourse.julialang.org/t/extend-distributed-pmap-to-mutliple-nodes-using-slurm/98263)

<div class="topic-metadata">

**Author:** [@psterzinger](https://discourse.julialang.org/u/psterzinger)\
**Replies:** 2\
**Last updated:** [May 3, 2023, 5:07pm UTC](https://discourse.julialang.org/t/extend-distributed-pmap-to-mutliple-nodes-using-slurm/98263 "2023-05-03T17:07:48Z")

</div>

I have a program that makes use of the parallelism offered by the Distributed’s pmap() function. I can run this on a cluster using one node and multiple cores. Now I would like to extend this to using multiple cores. Th…

---

## [MPIClusterManagers task split across nodes](https://discourse.julialang.org/t/mpiclustermanagers-task-split-across-nodes/97983)

<div class="topic-metadata">

**Author:** [@sparrowhawk](https://discourse.julialang.org/u/sparrowhawk)\
**Replies:** 2\
**Last updated:** [April 28, 2023, 2:02am UTC](https://discourse.julialang.org/t/mpiclustermanagers-task-split-across-nodes/97983 "2023-04-28T02:02:31Z")

</div>

Hi all, I’ve been using Julia on the cluster quite successfully with MPIClusterManagers.jl and MPI.jl with the MPIClusterManagers.start\_main\_loop(MPI\_TRANSPORT\_ALL) option and using native julia constructs after that su…

---

## [Interesting paper from Oak Ridge: Evaluating performance and portability of high-level programming models:](https://discourse.julialang.org/t/interesting-paper-from-oak-ridge-evaluating-performance-and-portability-of-high-level-programming-models/97451)

<div class="topic-metadata">

**Author:** [@Perrin\_Meyer](https://discourse.julialang.org/u/Perrin_Meyer)\
**Replies:** 1\
**Last updated:** [April 13, 2023, 9:43pm UTC](https://discourse.julialang.org/t/interesting-paper-from-oak-ridge-evaluating-performance-and-portability-of-high-level-programming-models/97451 "2023-04-13T21:43:18Z")

</div>

I noticed this interesting paper from Oak Ridge browsing arXiv: Evaluating performance and portability of high-level programming models: Julia, Python/Numba, and Kokkos on exascale nodes

---

## [ParallelStencil + ImplicitGlobalGrid with multiple GPUs](https://discourse.julialang.org/t/parallelstencil-implicitglobalgrid-with-multiple-gpus/96585)

<div class="topic-metadata">

**Author:** [@Teerthal](https://discourse.julialang.org/u/Teerthal)\
**Replies:** 8\
**Last updated:** [March 31, 2023, 7:39pm UTC](https://discourse.julialang.org/t/parallelstencil-implicitglobalgrid-with-multiple-gpus/96585 "2023-03-31T19:39:24Z")

</div>

Hello, I am currently starting out with GPU programming and working out a multiple GPU implementation of my Electroweak field theory simulation code. I have followed the examples on the ParallelStencil git for multi-GP…

---

## [Parallel merge in Cilk+ vs Julia](https://discourse.julialang.org/t/parallel-merge-in-cilk-vs-julia/34903)

<div class="topic-metadata">

**Author:** [@eaubanel](https://discourse.julialang.org/u/eaubanel)\
**Replies:** 18\
**Last updated:** [March 27, 2023, 10:02pm UTC](https://discourse.julialang.org/t/parallel-merge-in-cilk-vs-julia/34903 "2023-03-27T22:02:55Z")

</div>

As my first Julia program, I had some fun debugging a Julia implementation of a parallel merge, which I had already developed in Cilk+. I thought I’d make my first post by pointing the following out to warn others develo…

---

## [Use MPI with iJulia](https://discourse.julialang.org/t/use-mpi-with-ijulia/96319)

<div class="topic-metadata">

**Author:** [@smartalecH](https://discourse.julialang.org/u/smartalecH)\
**Replies:** 1\
**Last updated:** [March 19, 2023, 8:43pm UTC](https://discourse.julialang.org/t/use-mpi-with-ijulia/96319 "2023-03-19T20:43:01Z")

</div>

Is there an easy/recommended way to use iJulia with MPI? I see this has (somewhat) been asked in the past. But I’m wondering if anything has changed in the last few years (which is quite a bit of time for Julia…)

---

## [How to assign a device for each worker correctly in a multi gpu multi node scenario?](https://discourse.julialang.org/t/how-to-assign-a-device-for-each-worker-correctly-in-a-multi-gpu-multi-node-scenario/95715)

<div class="topic-metadata">

**Author:** [@hunKing](https://discourse.julialang.org/u/hunKing)\
**Replies:** 6\
**Last updated:** [March 9, 2023, 1:56am UTC](https://discourse.julialang.org/t/how-to-assign-a-device-for-each-worker-correctly-in-a-multi-gpu-multi-node-scenario/95715 "2023-03-09T01:56:48Z")

</div>

Hello folks, Does someone know how to correctly set the GPUs in each node correctly to individual workers in those nodes? For instance, imagine I have 2 nodes, with each having 2 GPUs. If I try to use the same procedure…

---

## [Julia seems to be running multithreaded when I don't want it to](https://discourse.julialang.org/t/julia-seems-to-be-running-multithreaded-when-i-dont-want-it-to/95391)

<div class="topic-metadata">

**Author:** [@nvenkov1](https://discourse.julialang.org/u/nvenkov1)\
**Replies:** 67\
**Last updated:** [March 2, 2023, 12:20pm UTC](https://discourse.julialang.org/t/julia-seems-to-be-running-multithreaded-when-i-dont-want-it-to/95391 "2023-03-02T12:20:08Z")

</div>

I coded an algorithm called CGS2, i.e., re-orthogonalized classical Gram-Schmidt, which I need to benchmark. I absolutely need my test to run on a single thread. To make sure it is the case, I write the following at the …

---

## [Data exchange between Julia sessions (same or different machines)](https://discourse.julialang.org/t/data-exchange-between-julia-sessions-same-or-different-machines/38157)

<div class="topic-metadata">

**Author:** [@martincornejo](https://discourse.julialang.org/u/martincornejo)\
**Replies:** 10\
**Last updated:** [February 24, 2023, 4:04pm UTC](https://discourse.julialang.org/t/data-exchange-between-julia-sessions-same-or-different-machines/38157 "2023-02-24T16:04:35Z")

</div>

Julia is great for implementing distributed computing. I was wondering how could it be possible to take a decentralized approach to this distributed computing. Instead of calling workers from a Julia (master) session (fo…

---

## [Unreachable host when adding remote workers](https://discourse.julialang.org/t/unreachable-host-when-adding-remote-workers/93591)

<div class="topic-metadata">

**Author:** [@fbanning](https://discourse.julialang.org/u/fbanning)\
**Replies:** 0\
**Last updated:** [January 26, 2023, 6:48pm UTC](https://discourse.julialang.org/t/unreachable-host-when-adding-remote-workers/93591 "2023-01-26T18:48:51Z")

</div>

Hiya, I’m trying to add some remote workers to my local julia process. Both host and remote run Julia 1.8.5 on Linux. I have passwordless access to the remote machine via SSH making it reachable through its alias withou…

---

## [MPI.jl memory issue in a for-loop](https://discourse.julialang.org/t/mpi-jl-memory-issue-in-a-for-loop/91721)

<div class="topic-metadata">

**Author:** [@Carol](https://discourse.julialang.org/u/Carol)\
**Replies:** 3\
**Last updated:** [January 2, 2023, 11:01am UTC](https://discourse.julialang.org/t/mpi-jl-memory-issue-in-a-for-loop/91721 "2023-01-02T11:01:37Z")

</div>

Hi, At first I want to say thank you for maintaining MPI.jl package. I am using MPI inside a for-loop (for iteration=1:5000). At each iteration, all ranks will send its data to rank0 using MPI.Gatherv!, then rank0 will…

---

## [Pkg.instantiate() error with multiple workers on 1 machine](https://discourse.julialang.org/t/pkg-instantiate-error-with-multiple-workers-on-1-machine/91022)

<div class="topic-metadata">

**Author:** [@jocklawrie](https://discourse.julialang.org/u/jocklawrie)\
**Replies:** 1\
**Last updated:** [December 1, 2022, 10:18pm UTC](https://discourse.julialang.org/t/pkg-instantiate-error-with-multiple-workers-on-1-machine/91022 "2022-12-01T22:18:53Z")

</div>

Hi there, This may well be user error, but thought I’d report it anyway in case there’s more to it. I’m running a simulation over multiple workers on the same machine. I get an error (see the image below) that indicate…

[Previous page](https://discourse.julialang.org/c/domain/parallel/34.md?page=1)

[Next page](https://discourse.julialang.org/c/domain/parallel/34.md?page=3)