Objectives

The primary objective of this tender is to develop, construct, and deploy a distributed network infrastructure that demonstrates a mesh computer network. This infrastructure will include hardware and software tools to enable open science research, facilitating communication and data processing across participating institutions.

Scope

The scope encompasses the end-to-end delivery of a self-contained, high-performance compute and data processing network integrated with existing infrastructure, including:

• Design and development of hardware and software components for a mesh network.
• Manufacturing and deployment of edge and centralized compute nodes.
• Integration with 5G base stations, edge data centers, and cloud platforms.
• Provision of ongoing maintenance and support post-deployment.
• Development of a comprehensive specification for the network platform, including hardware, software, and network considerations.

Technical Requirements

The technical specifications include:

• Hardware Components:
  • Edge Compute Nodes: Motherboard with PCIe Gen 4 x16, high-performance CPUs (Intel Xeon, AMD EPYC, or ARM SoCs), GPUs (NVIDIA A6000, AMD equivalent), 128 GB to 1 TB RAM, NVMe SSDs (1-2 TB), SATA SSDs (minimum 10 TB), 100 Gbps Ethernet, support for low-latency protocols, redundant power supplies, virtualization support.
  • Centralized Compute Nodes: High-performance processors (Intel Xeon, AMD EPYC), 256 GB to 2 TB RAM, hybrid storage (NVMe SSDs and HDDs), 100 Gbps Ethernet, Infiniband, GPU support (NVIDIA A100, H100), high availability features.
  • 5G NR hardware supporting various frequency bands, network slicing, integration with core network components, backhaul links, edge data centers with AI accelerators.
• Software Components:
  • Operating Systems: Ubuntu, CentOS, RHEL, Windows Server, specialized Linux distributions.
  • Hypervisors: KVM, VMware ESXi, Xen; containerization with Docker or Podman.
  • Network Virtualization: OpenStack, VMware vSphere, ETSI-compliant NFV MANO, SDN controllers (OpenFlow, proprietary solutions).
  • Edge Computing Platforms: Kubernetes, MEC platforms, integration with 5G network functions.
  • Cloud Management: Kubernetes, Prometheus, Grafana, ELK stack, CI/CD tools (Jenkins, GitLab), automation tools (Ansible, Terraform).
  • Application Hosting: Containerized applications, microservices, service mesh architectures (Istio, Linkerd).
  • Data Processing & AI/ML: TensorFlow, PyTorch, Hadoop, Spark, Cassandra, MongoDB, Kafka, Flink.
• Network & Infrastructure: Support for 5G network integration, edge data centers, secure cloud hybrid infrastructure, automation, and orchestration for scalability and flexibility.

Skills Requirements

The successful bidder must demonstrate:

• Proven experience in designing and deploying mesh and distributed network infrastructures.
• Capability to develop and manufacture custom hardware components, or effectively integrate off-the-shelf solutions.
• Expertise in 5G network hardware and software, including network slicing and core integration.
• Strong background in cloud-native technologies, virtualization, and container orchestration.
• Experience with AI/ML frameworks, big data processing, and real-time data analytics.
• Ability to provide ongoing maintenance, support, and skills transfer post-deployment.
• Knowledge of security protocols, DDoS protection, user authentication, and tamper-proofing devices.
• Capacity to implement automated management, monitoring, and fault recovery systems.