First updates for ISC25

Author: AndiH

.archive.mk

@@ -6,16 +6,16 @@
 # Changelog:
 # * Nov 2022: The archive is extracted again, then slides.pdf is removed if a patched slides-sc22.pdf is found (which includes an SC22 slide 0 title slide); and then repackaged
 .PHONY: all
-all: tut123-multi-gpu.tar.gz
+all: tut105-multi-gpu.tar.gz
 
-SOURCES=$(shell gfind . -maxdepth 1 -mindepth 1 -not -path "./.*" -not -name "tut123-multi-gpu.tar.gz" -printf '%P\n' | sort -h)
+SOURCES=$(shell gfind . -maxdepth 1 -mindepth 1 -not -path "./.*" -not -name "tut105-multi-gpu.tar.gz" -printf '%P\n' | sort -h)
 
-tut123-multi-gpu.tar.gz: $(shell find . -not -name "tut123-multi-gpu.tar.gz")
+tut105-multi-gpu.tar.gz: $(shell find . -not -name "tut105-multi-gpu.tar.gz")
 	sed -i '1 i***Please check GitHub repo for latest version of slides: https://github.com/FZJ-JSC/tutorial-multi-gpu/ ***\n' README.md
-	tar czf $@ --transform 's,^,SC24-tut123-Multi-GPU/,' --exclude=".*" $(SOURCES)
+	tar czf $@ --transform 's,^,ISC25-tut105-Multi-GPU/,' --exclude=".*" $(SOURCES)
 	tar xf $@
 	rm $@
-	find SC24-tut123-Multi-GPU/ -not -path './.*' -iname 'slides-*.pdf' -execdir rm slides.pdf \;
-	tar czf $@ SC24-tut123-Multi-GPU
-	rm -rf SC24-tut123-Multi-GPU
+	find ISC25-tut105-Multi-GPU/ -not -path './.*' -iname 'slides-*.pdf' -execdir rm slides.pdf \;
+	tar czf $@ ISC25-tut105-Multi-GPU
+	rm -rf ISC25-tut105-Multi-GPU
 	sed -i '1,2d' README.md

.etc/jsccourse-bashrc.sh

@@ -118,7 +118,7 @@ if [[ $- =~ "i" ]]; then
 
 	echo ""
 	echo "*******************************************************************************"
-	echo "       Welcome to the SC24 Tutorial on Multi-GPU Computing for Exascale!       "
+	echo "       Welcome to the ISC25 Tutorial on Multi-GPU Computing for Exascale!       "
 	# echo " A default call to get a batch system allocation is stored in \$JSC_ALLOC_CMD!"
 	# echo " Use it with \`eval \$JSC_ALLOC_CMD\`. The value of \$JSC_ALLOC_CMD is:"
 	# echo -n "  "

.gitignore

@@ -1,3 +1,3 @@
-tut123-multi-gpu.tar.gz
+tut105-multi-gpu.tar.gz
 *-sc24.pdf
 tut*

.zenodo.json

@@ -29,21 +29,21 @@
 
     "title": "Efficient Distributed GPU Programming for Exascale",
 
-    "publication_date": "2024-11-17",
+    "publication_date": "2025-06-13",
 
     "description": "<p>Over the past decade, GPUs became ubiquitous in HPC installations around the world, delivering the majority of performance of some of the largest supercomputers (e.g. Summit, Sierra, JUWELS Booster). This trend continues in the recently deployed and upcoming Pre-Exascale and Exascale systems (JUPITER, LUMI, Leonardo; El Capitan, Frontier, Aurora): GPUs are chosen as the core computing devices to enter this next era of HPC.To take advantage of future GPU-accelerated systems with tens of thousands of devices, application developers need to have the proper skills and tools to understand, manage, and optimize distributed GPU applications.In this tutorial, participants will learn techniques to efficiently program large-scale multi-GPU systems. While programming multiple GPUs with MPI is explained in detail, also advanced tuning techniques and complementing programming models like NCCL and NVSHMEM are presented. Tools for analysis are shown and used to motivate and implement performance optimizations. The tutorial teaches fundamental concepts that apply to GPU-accelerated systems in general, taking the NVIDIA platform as an example. It is a combination of lectures and hands-on exercises, using a development system for JUPITER (JEDI), for interactive learning and discovery.</p>",
 
-    "notes": "Slides and exercises of tutorial presented at SC24 (The International Conference for High Performance Computing, Networking, Storage, and Analysis 2024); https://sc24.conference-program.com/presentation/?id=tut123&sess=sess412",
+    "notes": "Slides and exercises of tutorial presented at ISC High Performance 2025; https://isc.app.swapcard.com/widget/event/isc-high-performance-2025/planning/UGxhbm5pbmdfMjU4MTc5Ng==",
 
     "access_right": "open",
 
-    "conference_title": "SC 2024",
-    "conference_acronym": "SC24",
-    "conference_dates": "17 November-22 November 2024",
-    "conference_place": "Atlanta, Georgia, USA",
-    "conference_url": "https://sc24.supercomputing.org/",
+    "conference_title": "ISC 2025",
+    "conference_acronym": "ISC25",
+    "conference_dates": "10 June-13 June 2025",
+    "conference_place": "Hamburg, Germany",
+    "conference_url": "https://www.isc-hpc.com/",
     "conference_session": "Tutorials",
-    "conference_session_part": "Day 1",
+    "conference_session_part": "Afternoon",
 
     "upload_type": "lesson"
 }

CITATION.cff

@@ -48,5 +48,5 @@ keywords:
   - NVSHMEM
   - Distributed Programming
 license: MIT
-version: '7.0-sc24'
-date-released: '2024-11-17'
+version: '8.0-isc25'
+date-released: '2025-06-13'

README.md

@@ -1,15 +1,15 @@
-# SC24 Tutorial: Efficient Distributed GPU Programming for Exascale
+# ISC25 Tutorial: Efficient Distributed GPU Programming for Exascale
 
 [![DOI](https://zenodo.org/badge/409504932.svg)](https://zenodo.org/badge/latestdoi/409504932)
 
 
-Repository with talks and exercises of our Efficient GPU Programming for Exascale tutorial, to be held at [SC24](https://sc24.conference-program.com/presentation/?id=tut123&sess=sess412).
+Repository with talks and exercises of our Efficient GPU Programming for Exascale tutorial, to be held at [ISC25](https://isc.app.swapcard.com/widget/event/isc-high-performance-2025/planning/UGxhbm5pbmdfMjU4MTc5Ng==).
 
 ## Coordinates
 
-* Date: 17 November 2024
-* Occasion: SC24 Tutorial
-* Tutors: Simon Garcia de Gonzalo (SNL), Andreas Herten (JSC), Markus Hrywniak (NVIDIA), Jiri Kraus (NVIDIA), Lena Oden (Uni Hagen)
+* Date: 13 June 2025
+* Occasion: ISC25 Tutorial
+* Tutors: Simon Garcia de Gonzalo (SNL), Andreas Herten (JSC), Lena Oden (Uni Hagen), with support by Markus Hrywniak (NVIDIA) and Jiri Kraus (NVIDIA)
 
 
 ## Setup
@@ -20,21 +20,21 @@ Walk-through:
 
 * Sign up at JuDoor
 * Open Jupyter JSC: https://jupyter-jsc.fz-juelich.de
-* Create new Jupyter instance on JEDI, using training2446 account, on **LoginNode**
-* Source course environment: `source $PROJECT_training2446/env.sh`
+* Create new Jupyter instance on JEDI, using training25XX account, on **LoginNode**
+* Source course environment: `source $PROJECT_training25XX/env.sh`
 * Sync material: `jsc-material-sync`
 * Locally install NVIDIA Nsight Systems: https://developer.nvidia.com/nsight-systems
 
-Curriculum:
+Curriculum (Note: square-bracketed sessions are skipped at ISC25 because only ½ day was allocated to the tutorial):
 
 1. Lecture: Tutorial Overview, Introduction to System + Onboarding *Andreas*
 2. Lecture: MPI-Distributed Computing with GPUs *Simon*
 3. Hands-on: Multi-GPU Parallelization
-4. Lecture: Performance / Debugging Tools *Markus*
-5. Lecture: Optimization Techniques for Multi-GPU Applications *Simon*
+4. [Lecture: Performance / Debugging Tools]
+5. Lecture: Optimization Techniques for Multi-GPU Applications *Lena*
 6. Hands-on: Overlap Communication and Computation with MPI
-7. Lecture: Overview of NCCL and NVSHMEN in MPI *Lena*
-8. Hands-on: Using NCCL and NVSHMEM
-9. Lecture: Device-initiated Communication with NVSHMEM *Jiri*
-10. Hands-on: Using Device-Initiated Communication with NVSHMEM
+7. [Lecture: Overview of NCCL and NVSHMEN in MPI]
+8. [Hands-on: Using NCCL and NVSHMEM]
+9. [Lecture: Device-initiated Communication with NVSHMEM]
+10. [Hands-on: Using Device-Initiated Communication with NVSHMEM]
 11. Lecture: Conclusion and Outline of Advanced Topics *Andreas*