Skip to content

Commit cbda7c8

Browse files
ranimandepudiranimandepudi
authored
Merge pull request #909 from ArmDeveloperEcosystem/main
Added Nutanix, NVIDIA Cuda & few CSPs
2 parents dcbfec1 + 0579d5f commit cbda7c8

39 files changed

Lines changed: 1183 additions & 0 deletions

‎content/linux/opensource_packages/amazon-vpc-cni-k8s.md‎

Lines changed: 31 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,31 @@
1+
---
2+
name: Nutanix Amazon-vpc-cni-k8s
3+
category: Networking
4+
description: Amazon-vpc-cni-k8s is a Kubernetes Container Network Interface (CNI) plugin that enables pod networking by directly integrating Kubernetes pods with AWS VPC networking using Elastic Network Interfaces (ENIs).
5+
download_url: https://github.com/nutanix/amazon-vpc-cni-k8s/tags
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 1.4.0
9+
release_date: 2019/04/16
10+
11+
12+
optional_info:
13+
homepage_url: https://github.com/nutanix/amazon-vpc-cni-k8s
14+
support_caveats:
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://github.com/nutanix/amazon-vpc-cni-k8s#building
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number:
22+
release_date:
23+
reference_content:
24+
rationale:
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum: https://github.com/nutanix/amazon-vpc-cni-k8s/blob/master/CHANGELOG.md#v140
28+
release_notes__recommended_minimum:
29+
other_info:
30+
31+
---

‎content/linux/opensource_packages/ampere-ai-text-to-sql.md‎

Lines changed: 32 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,32 @@
1+
---
2+
name: Ampere AI Text-to-SQL
3+
category: AI/ML
4+
description: Ampere AI Text-to-SQL is a reference AI application that enables users to query databases using natural language by integrating large language models with Open WebUI and LlamaIndex, optimized for deployment on Ampere CPU-based systems using Docker.
5+
download_url: https://github.com/orgs/AmpereComputingAI/packages/container/package/ampere-ai-text2sql
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 0.1
9+
release_date: 2025/12/06
10+
11+
12+
optional_info:
13+
homepage_url: https://github.com/AmpereComputingAI/ampere-ai-text2sql
14+
support_caveats:
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://github.com/AmpereComputingAI/ampere-ai-text2sql#prerequisites
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number:
22+
release_date:
23+
reference_content:
24+
rationale:
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum:
28+
release_notes__recommended_minimum:
29+
other_info: Although the repository does not explicitly state Arm64 support, the prerequisite requirement for Ampere Altra or AmpereOne systems, both Arm64-only processors, indicates that the package is designed for and supports Linux Arm64 starting from its initial release (v0.1).
30+
31+
---
32+

‎content/linux/opensource_packages/ampere-optimised-ollama.md‎

Lines changed: 32 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,32 @@
1+
---
2+
name: Ampere Optimized Ollama
3+
category: AI/ML
4+
description: Ampere Optimized Ollama is an optimized build of Ollama designed to efficiently run and serve large language models on Ampere CPUs, delivering improved inference performance through architecture-specific optimizations and support for custom quantized models.
5+
download_url: https://hub.docker.com/r/amperecomputingai/ollama/tags
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 1.0.0-ol9
9+
release_date: 2025/07/26
10+
11+
12+
optional_info:
13+
homepage_url: https://hub.docker.com/r/amperecomputingai/ollama
14+
support_caveats: The [Docker image for Ollama by Ampere Computing](https://hub.docker.com/r/amperecomputingai/ollama/tags) can be run on bare metal Ampere CPUs and Ampere-based VMs available in the cloud.
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://hub.docker.com/r/amperecomputingai/ollama#running
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number:
22+
release_date:
23+
reference_content:
24+
rationale:
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum:
28+
release_notes__recommended_minimum:
29+
other_info: The release notes specific to Linux/Arm64 support isn't available. Linux/Arm64 Ollama docker images by Ampere Computing are available from version 1.0.0-ol9 onwards.
30+
31+
---
32+

‎content/linux/opensource_packages/ampere-optimized-llama.md‎

Lines changed: 31 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,31 @@
1+
---
2+
name: Ampere Optimized Llama.cpp
3+
category: AI/ML
4+
description: Ampere Optimized llama.cpp is an optimized build of llama.cpp designed to run GGUF large language models efficiently on Ampere CPUs, providing improved inference performance through architecture-specific optimizations.
5+
download_url: https://github.com/AmpereComputingAI/llama.cpp/releases
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 1.2.0
9+
release_date: 2024/05/16
10+
11+
12+
optional_info:
13+
homepage_url: https://github.com/AmpereComputingAI/llama.cpp
14+
support_caveats: The [Docker image for llama.cpp by Ampere Computing](https://hub.docker.com/r/amperecomputingai/llama.cpp/tags) can be run on bare metal Ampere CPUs and Ampere-based VMs available in the cloud.
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://github.com/AmpereComputingAI/llama.cpp#starting-container
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number:
22+
release_date:
23+
reference_content:
24+
rationale:
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum:
28+
release_notes__recommended_minimum:
29+
other_info: The release notes specific to Linux/Arm64 support isn't available. Linux/Arm64/v8 Llama.cpp docker images by Ampere Computing are available from version 1.2.0 onwards.
30+
31+
---

‎content/linux/opensource_packages/ampere-optimized-onnx.md‎

Lines changed: 31 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,31 @@
1+
---
2+
name: Ampere Optimized ONNX Runtime
3+
category: AI/ML
4+
description: Ampere Optimized ONNX Runtime is an inference acceleration backend for ONNX Runtime that enhances deep learning performance on Ampere Altra ARM CPUs using model optimizations, vectorized compute kernels, and multi-threaded execution without requiring model changes.
5+
download_url: https://hub.docker.com/r/amperecomputingai/onnxruntime/tags
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 1.4.0
9+
release_date: 2022/12/02
10+
11+
12+
optional_info:
13+
homepage_url: https://amperecomputing.com/developers/power-your-ai
14+
support_caveats: Ampere Optimized ONNX Runtime targets the Ampere Altra family of Arm64 processors. While Arm64 compatibility may predate 1.4.0, official packaged distributions (Docker images) are available starting from version 1.4.0.
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://amperecomputing.com/assets/Ampere_Optimized_ONNXRuntime_Documentation_v1_8_0_9646259707.pdf
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number:
22+
release_date:
23+
reference_content:
24+
rationale:
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum:
28+
release_notes__recommended_minimum:
29+
other_info: The release notes for the initial Linux/Arm64 support isn't available. Docker images for Aarch64 are available from version 1.4.0 onwards, which officially justifies the Arm support.
30+
31+
---

‎content/linux/opensource_packages/ampere-optimized-pytorch.md‎

Lines changed: 31 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,31 @@
1+
---
2+
name: Ampere Optimized PyTorch
3+
category: AI/ML
4+
description: Ampere Optimized PyTorch is a deep learning acceleration backend that enhances PyTorch inference performance on Ampere Altra ARM CPUs using optimized kernels, vectorization, and multi-threading for improved latency and throughput without requiring model changes.
5+
download_url: https://hub.docker.com/r/amperecomputingai/pytorch/tags
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 1.4.0
9+
release_date: 2022/12/02
10+
11+
12+
optional_info:
13+
homepage_url: https://amperecomputing.com/developers/power-your-ai
14+
support_caveats: Ampere Optimized PyTorch targets the Ampere Altra family of Arm64 processors and integrates the Ampere AI backend for accelerated inference. While Arm64 compatibility may predate 1.4.0, official packaged distributions (Docker images) are available starting from version 1.4.0.
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://amperecomputing.com/assets/Ampere_Optimized_PyTorch_Documentation_v1_8_0_e61744b94b.pdf
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number:
22+
release_date:
23+
reference_content:
24+
rationale:
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum:
28+
release_notes__recommended_minimum:
29+
other_info: The release notes for the initial Linux/Arm64 support isn't available. Docker images for Aarch64 are available from version 1.4.0 onwards, which officially justifies the Arm support.
30+
31+
---

‎content/linux/opensource_packages/ampere-optimized-tensorflow.md‎

Lines changed: 31 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,31 @@
1+
---
2+
name: Ampere Optimized TensorFlow
3+
category: AI/ML
4+
description: Ampere Optimized TensorFlow is an inference acceleration backend that improves TensorFlow performance on Ampere Altra ARM CPUs using optimized compute kernels and multi-threading, enabling faster execution without requiring changes to existing TensorFlow models.
5+
download_url: https://hub.docker.com/r/amperecomputingai/tensorflow/tags
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 1.4.0
9+
release_date: 2022/12/01
10+
11+
12+
optional_info:
13+
homepage_url: https://amperecomputing.com/developers/power-your-ai
14+
support_caveats: Ampere Optimized TensorFlow targets the Ampere Altra family of Arm64 processors. While Arm64 compatibility may predate 1.4.0, official packaged distributions (Docker images) are available starting from version 1.4.0.
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://amperecomputing.com/assets/Ampere_Optimized_Tensorflow_Documentation_v_1_8_0_8ab3f5a054.pdf
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number:
22+
release_date:
23+
reference_content:
24+
rationale:
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum:
28+
release_notes__recommended_minimum:
29+
other_info: The release notes for the initial Linux/Arm64 support isn't available. Docker images for Aarch64 are available from version 1.4.0 onwards, which officially justifies the Arm support.
30+
31+
---

‎content/linux/opensource_packages/bionemo-framework.md‎

Lines changed: 31 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,31 @@
1+
---
2+
name: NVIDIA BioNeMo-Framework
3+
category: AI/ML
4+
description: BioNeMo is an NVIDIA software ecosystem for building, training, fine-tuning, and deploying AI models for life sciences, providing optimized biomolecular models, workflows, and production-ready inference microservices for biological and therapeutic applications.
5+
download_url: https://catalog.ngc.nvidia.com/orgs/nvidia/teams/clara/containers/bionemo-framework
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 2.2
9+
release_date: 2024/12/19
10+
11+
12+
optional_info:
13+
homepage_url: https://nvidia.github.io/bionemo-framework/
14+
support_caveats: BioNeMo Framework v2.2+ provides Arm64 container images (e.g., for GH200 platform). Documentation for BioNeMo lists supported GPUs with Compute Capability ≥8.0 (H100, L4, L40, A100, A40, A30, A10, A16, A2, RTX 6000/A6000/A5000/A4000) on supported hosts. Arm64 GPU validation details are not separately published. Users should confirm compatibility with their CUDA driver and GPU architecture. Please refer [this](https://nvidia.github.io/bionemo-framework/main/getting-started/pre-reqs/).
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://nvidia.github.io/bionemo-framework/main/getting-started/pre-reqs/
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number: 2.6
22+
release_date: 2025/04/30
23+
reference_content: https://github.com/NVIDIA/bionemo-framework/releases/tag/v2.6
24+
rationale: In this release, docker build was fixed for Arm and Arm compatible containers were released.
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum: https://github.com/NVIDIA/bionemo-framework/releases/tag/v2.2
28+
release_notes__recommended_minimum:
29+
other_info:
30+
31+
---

‎content/linux/opensource_packages/clara-viz.md‎

Lines changed: 31 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,31 @@
1+
---
2+
name: Clara-Viz
3+
category: Video
4+
description: NVIDIA Clara Viz is a CUDA-accelerated medical imaging visualization platform that enables high-performance 2D and 3D volumetric rendering and multi-resolution image viewing through GPU-based ray tracing, with Python APIs and interactive notebook widgets.
5+
download_url: https://github.com/NVIDIA/clara-viz/releases
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 0.3.0
9+
release_date: 2023/06/22
10+
11+
12+
optional_info:
13+
homepage_url: https://docs.nvidia.com/clara-viz/index.html
14+
support_caveats: Supported NVIDIA GPUs are Pascal or newer (including Pascal, Volta, Turing and Ampere families), and driver 450.36.06+ is required.
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://docs.nvidia.com/clara-viz/installation.html
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number:
22+
release_date:
23+
reference_content:
24+
rationale:
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum: https://github.com/NVIDIA/clara-viz/releases/tag/v0.3.0
28+
release_notes__recommended_minimum:
29+
other_info:
30+
31+
---

‎content/linux/opensource_packages/cloud-native-stack.md‎

Lines changed: 31 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,31 @@
1+
---
2+
name: NVIDIA Cloud Native Stack
3+
category: Containers and Orchestration
4+
description: NVIDIA Cloud Native Stack (CNS) is a reference cloud-native software stack for running GPU-accelerated workloads on Kubernetes, combining Ubuntu/RHEL, Kubernetes, Helm, and NVIDIA GPU and Network Operators to deploy and manage NVIDIA GPUs in containerized environments.
5+
download_url: https://github.com/NVIDIA/cloud-native-stack/releases
6+
works_on_arm: true
7+
supported_minimum_version:
8+
version_number: 8.0
9+
release_date: 2022/10/14
10+
11+
12+
optional_info:
13+
homepage_url: https://github.com/NVIDIA/cloud-native-stack
14+
support_caveats: NVIDIA Cloud Native Stack provides official support for server-class Linux Arm64 (Aarch64) platforms starting with version 8.0 (October 2022), where “NVIDIA Certified Server (x86 & Arm64)” is explicitly listed. Feature availability (e.g., Network Operator) may vary by release and is not always at parity with x86. Kindly refer [Component Matrix](https://github.com/NVIDIA/cloud-native-stack#component-matrix) for more details.
15+
alternative_options:
16+
getting_started_resources:
17+
official_docs: https://github.com/NVIDIA/cloud-native-stack#getting-started
18+
arm_content:
19+
partner_content:
20+
arm_recommended_minimum_version:
21+
version_number:
22+
release_date:
23+
reference_content:
24+
rationale:
25+
26+
optional_hidden_info:
27+
release_notes__supported_minimum: https://github.com/NVIDIA/cloud-native-stack/tree/v23.8.0#nvidia-cloud-native-stack-component-matrix-1
28+
release_notes__recommended_minimum:
29+
other_info: Initial official Linux/Arm64 (Aarch64) support for NVIDIA Cloud Native Stack appears in version 8.0 (Oct 2022), based on the first explicit inclusion of “NVIDIA Certified Server (x86 & Arm64)” in the component matrix.
30+
31+
---

0 commit comments

Comments
 (0)