---
title: "GPUs Cluster (Panther)"
canonical: "https://qdocs.qcri.org/page/gpus-cluster-panther/"
format: markdown
---
# **NVIDIA Bright Cluster Manager**

Is the core software stack of the QCRI Panther Cluster, offering fast deployment and end-to-end administration, management and monitoring for the heterogeneous high-performance computing and AI/GPU-enabled server clusters in the data center, and in multi/hybrid-cloud environments. It supports CPU-based and NVIDIA GPU-accelerated systems, and enables orchestration with Kubernetes.

Other features of Bright Cluster Manager include:

-   **Modern deep learning environment** — Provides everything that you require to quickly set up a complete cluster-ready and easy-to-manage machine learning environment for data scientists.  
    
-   **Choice of machine learning libraries and frameworks** — Includes the most popular machine learning libraries and frameworks such as NVIDIA cuDNN, CUB, CUDA, TensorRT, DyNet, Fastai, JupyterHub, NCCL2, MXNet, PyTorch, Chainer, CNTK, Horovod, Keras, Opencv3, Protobuf3, Torch, TensorFlow, Theano, and XGBoost  
    
-   **Simple deployment process** — Eases installation and configuration. After you answer a few questions about your cluster, the Bright software is installed, and the necessary configuration files are created.  
    
-   **Flexible workload scheduling** — The default workload scheduling of jobs is being done through Slurm

![](/images/c1f7c0ef-a75d-44e3-bee9-475c41c34509)

### **NVIDIA Bright Cluster Manager**

NVIDIA Bright Cluster Manager is the core software stack of the QCRI Panther Cluster, offering fast deployment and end-to-end administration for high-performance computing and AI/GPU-enabled server clusters. It supports CPU-based and NVIDIA GPU-accelerated systems.

[Learn More](https://qdocs.qcri.org/space/services/38667294/Panther+GPU+Cluster)

![](/images/cb25ff8d-0267-45ff-b0b7-759234486015)

### **Features of Bright Cluster Manager**

Some features of Bright Cluster Manager include a modern deep learning environment, choice of machine learning libraries and frameworks, simple deployment process, and flexible workload scheduling through Slurm scheduling.

[Learn More](https://qdocs.qcri.org/space/services/38667294/Panther+GPU+Cluster)

![](/images/c1f7c0ef-a75d-44e3-bee9-475c41c34509)

### **Benefits of Bright Cluster Manager**

The benefits of using Bright Cluster Manager include quick setup of a complete machine learning environment, support for various popular machine learning libraries, easy installation and configuration process, and flexible workload scheduling.

[https://qdocs.qcri.org/space/services/38667294/Panther+GPU+Cluster](https://qdocs.qcri.org/space/services/38667294/Panther+GPU+Cluster)

> Dynamic module: Pages (live content)

## **Links - Internal References**

**CQL:** `type = page and lastModified <= endOfYear() order by lastModified desc`

## Children

- Panther