From f5a848087345ae39d718cf28c781787121916f70 Mon Sep 17 00:00:00 2001 From: Samuel Gozel Date: Fri, 4 Sep 2026 15:37:17 +0200 Subject: [PATCH 1/4] add doc for beverin cluster --- docs/access/firecrest.md | 2 +- docs/alps/hardware.md | 15 ++++++ docs/clusters/beverin.md | 113 +++++++++++++++++++++++++++++++++++++++ docs/clusters/index.md | 8 +++ zensical.toml | 2 + 5 files changed, 139 insertions(+), 1 deletion(-) create mode 100644 docs/clusters/beverin.md diff --git a/docs/access/firecrest.md b/docs/access/firecrest.md index cfaa3cafa..becb3a9d9 100644 --- a/docs/access/firecrest.md +++ b/docs/access/firecrest.md @@ -35,7 +35,7 @@ FirecREST is available for all three major [Alps platforms][ref-alps-platforms], | [HPC Platform][ref-platform-hpcp] | https://api.cscs.ch/hpc/firecrest/v2 | [Daint][ref-cluster-daint], [Eiger][ref-cluster-eiger] | | [ML Platform][ref-platform-mlp] | https://api.cscs.ch/ml/firecrest/v2 | [Bristen][ref-cluster-bristen], [Clariden][ref-cluster-clariden] | | [C&W Platform][ref-platform-cwp] | https://api.cscs.ch/cw/firecrest/v2 | [Santis][ref-cluster-santis] | -| Beverin | https://api.cscs.ch/beverin/firecrest/v2 | Beverin | +| [Beverin][ref-cluster-beverin] | https://api.cscs.ch/beverin/firecrest/v2 | [Beverin][ref-cluster-beverin] | ## Accessing FirecREST diff --git a/docs/alps/hardware.md b/docs/alps/hardware.md index 5a3d442ea..08c06d0f3 100644 --- a/docs/alps/hardware.md +++ b/docs/alps/hardware.md @@ -139,6 +139,14 @@ The Grizzly Peak blades contain two nodes, where each node has: Bard Peak +1 node: + +* 1 socket AMD Epyc 7A53 "Trento" 64-Core processor with 2 threads per core +* 512 GB DDR4 memory +* 4 AMD MI250X liquid-cooled OAMs for a total of 8 Graphics Compute Dies (GCDs) and 64 GB HBM2e memory per die +* 4 NICS -- one per OAM + + [](){#ref-alps-mi300-node} ### AMD MI300A GPU Nodes @@ -147,3 +155,10 @@ Bard Peak !!! todo Parry Peak + +1 node: + +* 4 MI300A Accelerated Processing Units (APUs) + * 24 Zen4 CPU cores per APU with 2 threads per core, for a total of 192 threads per node +* 512 GB unified physical HBM3e memory +* 4 NICS diff --git a/docs/clusters/beverin.md b/docs/clusters/beverin.md new file mode 100644 index 000000000..f4aa8f6ac --- /dev/null +++ b/docs/clusters/beverin.md @@ -0,0 +1,113 @@ +[](){#ref-cluster-beverin} +# Beverin + +Beverin is an AMD-based GPU cluster for the purpose of testing and porting applications. + +## Cluster Specification + +### Login node + +Beverin has one [MI250X][ref-alps-mi200-node] login node. + +### Compute Nodes + +Beverin consists of 13 [MI250X][ref-alps-mi200-node] and 128 [MI300A][ref-alps-mi300-node] compute nodes. + +| node type | number of nodes | total CPU sockets | total GPUs | +|-----------|--------| ----------------- | ---------- | +| [mi200][ref-alps-mi200-node] | 13 | 13 | 104 | +| [mi300][ref-alps-mi300-node] | 128 | 512 | 512 | + +### Storage and file systems + +Beverin uses the [HPCP filesystems and storage policies][ref-hpcp-storage]. + +## Getting started + +### Logging into Beverin + +To connect to Beverin via SSH, first refer to the [ssh guide][ref-ssh]. + +!!! example "`~/.ssh/config`" + Add the following to your [SSH configuration][ref-ssh-config] to enable you to directly connect to Beverin using `ssh beverin`. + ``` + Host beverin + HostName beverin.vc.cscs.ch + ProxyJump ela + User cscsusername + IdentityFile ~/.ssh/cscs-key + IdentitiesOnly yes + ``` + +### Software + +[](){#ref-cluster-beverin-uenv} + +Beverin provides uenv to deliver programming environments and application software. Please refer to the [uenv documentation][ref-uenv] for detailed information on how to use the uenv tools on the system. + +
+ +- :fontawesome-solid-layer-group: __Climate and Weather Applications__ + + Provide software stacks for climate and weather workflows on Beverin. + + * [ICON][ref-software-icon] + +
+ +
+ +- :fontawesome-solid-layer-group: __Scientific Applications__ + + Provide scientific applications. + + * [Quantumespresso][ref-uenv-quantumespresso] + +
+ +
+ +- :fontawesome-solid-layer-group: __Programming Environments__ + + Provide compilers, MPI, Python, common libraries and tools used to build your own applications. + + * [prgenv-gnu][ref-uenv-prgenv-gnu] + * [linalg][ref-uenv-linalg] + +
+ +
+ +- :fontawesome-solid-layer-group: __Tools__ + + Provide tools like + + * [Linaro Forge][ref-uenv-linaro] +
+ +[](){#ref-cluster-beverin-containers} +#### Containers + +Beverin supports container workloads using the [Container Engine][ref-container-engine]. + +To build images, see the [guide to building container images on Alps][ref-build-containers]. + +## Running Jobs on Beverin + +### Slurm + +Beverin uses [Slurm][ref-slurm] as the workload manager, which is used to launch and monitor distributed workloads. + +There are two Slurm partitions on the system, corresponding to the two node types: + +* `mi200` (default) +* `mi300` + +| name | nodes | max nodes per job | time limit | +| -- | -- | -- | -- | +| `mi200` | 13 | 13 | 24 hours | +| `mi300` | 128 | 128 | 24 hours | + +### FirecREST + +Beverin can also be accessed using [FirecREST][ref-firecrest] at the `https://api.cscs.ch/ml/firecrest/v2` API endpoint. diff --git a/docs/clusters/index.md b/docs/clusters/index.md index dc6de5f6b..f1728c90b 100644 --- a/docs/clusters/index.md +++ b/docs/clusters/index.md @@ -45,6 +45,14 @@ The following clusters are part of the platforms that are fully operated by CSCS ## Other systems +
+- :fontawesome-solid-mountain: __Testing, Porting and Development__ + + Beverin is a medium-sized system used for testing, developing and porting applications to AMD GPUs. + + [:octicons-arrow-right-24: Beverin][ref-cluster-beverin] +
+
- :fontawesome-solid-mountain: __Porting and Development__ diff --git a/zensical.toml b/zensical.toml index f73416429..bdc7fe745 100644 --- a/zensical.toml +++ b/zensical.toml @@ -22,6 +22,7 @@ nav = [ {"Clusters" = [ "clusters/index.md", {"Besso" = "clusters/besso.md"}, + {"Beverin" = "clusters/beverin.md"}, {"Bristen" = "clusters/bristen.md"}, {"Clariden" = "clusters/clariden.md"}, {"Daint" = "clusters/daint.md"}, @@ -40,6 +41,7 @@ nav = [ "platforms/hpcp/index.md", {"Daint" = "clusters/daint.md"}, {"Eiger" = "clusters/eiger.md"}, + {"Beverin" = "clusters/beverin.md"}, ]}, {"Climate and Weather Platform" = [ "platforms/cwp/index.md", From 9c5d2e4094e93b8422b34b620c67db940387f7ce Mon Sep 17 00:00:00 2001 From: Samuel Gozel Date: Fri, 4 Sep 2026 17:25:10 +0200 Subject: [PATCH 2/4] Add details on HPE Cray blades + improve structure --- .github/actions/spelling/allow.txt | 3 +++ docs/alps/hardware.md | 15 +++------------ docs/clusters/beverin.md | 3 +++ docs/clusters/index.md | 11 +++-------- zensical.toml | 1 - 5 files changed, 12 insertions(+), 21 deletions(-) diff --git a/.github/actions/spelling/allow.txt b/.github/actions/spelling/allow.txt index 23ad9bb43..f7bef873a 100644 --- a/.github/actions/spelling/allow.txt +++ b/.github/actions/spelling/allow.txt @@ -3,6 +3,7 @@ AMD AMPI Alpstein Apertus +APUs autoscaling auditability balfrin @@ -50,6 +51,7 @@ Fock Foket GAPW GBit +GCDs GGA GPFS GPG @@ -108,6 +110,7 @@ NVSHMEM NVLINK Nordend netcdf +OAMs OpenFabrics OAuth OIDC diff --git a/docs/alps/hardware.md b/docs/alps/hardware.md index 08c06d0f3..74e348393 100644 --- a/docs/alps/hardware.md +++ b/docs/alps/hardware.md @@ -135,28 +135,19 @@ The Grizzly Peak blades contain two nodes, where each node has: [](){#ref-alps-mi200-node} ### AMD MI250x GPU Nodes -!!! todo - -Bard Peak - -1 node: +Each HPE Cray EX235A (code-name Bard Peak) blade contains two nodes, where each node is made of: * 1 socket AMD Epyc 7A53 "Trento" 64-Core processor with 2 threads per core * 512 GB DDR4 memory * 4 AMD MI250X liquid-cooled OAMs for a total of 8 Graphics Compute Dies (GCDs) and 64 GB HBM2e memory per die -* 4 NICS -- one per OAM - +* 4 NICs -- one per OAM [](){#ref-alps-mi300-node} ### AMD MI300A GPU Nodes ![](../images/alps/mi300-schematic.svg) -!!! todo - -Parry Peak - -1 node: +Each HPE Cray EX255A (code-name Parry Peak) blade contains two nodes, where each node is made of: * 4 MI300A Accelerated Processing Units (APUs) * 24 Zen4 CPU cores per APU with 2 threads per core, for a total of 192 threads per node diff --git a/docs/clusters/beverin.md b/docs/clusters/beverin.md index f4aa8f6ac..f312b8a68 100644 --- a/docs/clusters/beverin.md +++ b/docs/clusters/beverin.md @@ -18,6 +18,8 @@ Beverin consists of 13 [MI250X][ref-alps-mi200-node] and 128 [MI300A][ref-alps-m | [mi200][ref-alps-mi200-node] | 13 | 13 | 104 | | [mi300][ref-alps-mi300-node] | 128 | 512 | 512 | +See [Running jobs on Beverin][ref-cluster-beverin-running-jobs] for details on the associated slurm partitions. + ### Storage and file systems Beverin uses the [HPCP filesystems and storage policies][ref-hpcp-storage]. @@ -92,6 +94,7 @@ Beverin supports container workloads using the [Container Engine][ref-container- To build images, see the [guide to building container images on Alps][ref-build-containers]. +[](){#ref-cluster-beverin-running-jobs} ## Running Jobs on Beverin ### Slurm diff --git a/docs/clusters/index.md b/docs/clusters/index.md index f1728c90b..e073f1b5b 100644 --- a/docs/clusters/index.md +++ b/docs/clusters/index.md @@ -51,14 +51,9 @@ The following clusters are part of the platforms that are fully operated by CSCS Beverin is a medium-sized system used for testing, developing and porting applications to AMD GPUs. [:octicons-arrow-right-24: Beverin][ref-cluster-beverin] -
- -
-- :fontawesome-solid-mountain: __Porting and Development__ - + Besso is a small system used by some partners for development and porting with AMD and NVIDIA GPUs. - + [:octicons-arrow-right-24: Besso][ref-cluster-besso] -
- + diff --git a/zensical.toml b/zensical.toml index bdc7fe745..9d15415c8 100644 --- a/zensical.toml +++ b/zensical.toml @@ -41,7 +41,6 @@ nav = [ "platforms/hpcp/index.md", {"Daint" = "clusters/daint.md"}, {"Eiger" = "clusters/eiger.md"}, - {"Beverin" = "clusters/beverin.md"}, ]}, {"Climate and Weather Platform" = [ "platforms/cwp/index.md", From d206d79a69fc8c1f5a3f800894b08103b53d6b55 Mon Sep 17 00:00:00 2001 From: Samuel Gozel Date: Fri, 4 Sep 2026 17:28:46 +0200 Subject: [PATCH 3/4] fix spell-check issue --- docs/alps/hardware.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/alps/hardware.md b/docs/alps/hardware.md index 74e348393..ff700a237 100644 --- a/docs/alps/hardware.md +++ b/docs/alps/hardware.md @@ -152,4 +152,4 @@ Each HPE Cray EX255A (code-name Parry Peak) blade contains two nodes, where each * 4 MI300A Accelerated Processing Units (APUs) * 24 Zen4 CPU cores per APU with 2 threads per core, for a total of 192 threads per node * 512 GB unified physical HBM3e memory -* 4 NICS +* 4 NICs -- one per MI300A APU From 2f798e50b837d8c1b509c74c8a2b868a5b5485df Mon Sep 17 00:00:00 2001 From: sgozel Date: Tue, 8 Sep 2026 18:27:05 +0200 Subject: [PATCH 4/4] update Beverin software section --- docs/clusters/beverin.md | 27 ++++++++++++++++----------- 1 file changed, 16 insertions(+), 11 deletions(-) diff --git a/docs/clusters/beverin.md b/docs/clusters/beverin.md index f312b8a68..f047da5cb 100644 --- a/docs/clusters/beverin.md +++ b/docs/clusters/beverin.md @@ -49,6 +49,19 @@ Beverin provides uenv to deliver programming environments and application softwa
+- :fontawesome-solid-layer-group: __Programming Environments__ + + Provide compilers, MPI, Python, common libraries and tools used to build your own applications. + + * [prgenv-gnu][ref-uenv-prgenv-gnu] + * [linalg][ref-uenv-linalg] + +
+ +In addition to the base `GNU` and `linalg` programming environments, a few other dedicated environments are provided on a best effort basis: + +
+ - :fontawesome-solid-layer-group: __Climate and Weather Applications__ Provide software stacks for climate and weather workflows on Beverin. @@ -69,17 +82,6 @@ Beverin provides uenv to deliver programming environments and application softwa
-- :fontawesome-solid-layer-group: __Programming Environments__ - - Provide compilers, MPI, Python, common libraries and tools used to build your own applications. - - * [prgenv-gnu][ref-uenv-prgenv-gnu] - * [linalg][ref-uenv-linalg] - -
- -
- - :fontawesome-solid-layer-group: __Tools__ Provide tools like @@ -87,6 +89,9 @@ Beverin provides uenv to deliver programming environments and application softwa * [Linaro Forge][ref-uenv-linaro]
+See the [uenv quick start guide][ref-uenv-quickstart] to find all programming environments available on Beverin. + + [](){#ref-cluster-beverin-containers} #### Containers