HPE ProLiant DL380 Gen11 NVIDIA L40S Configure-to-Order Server

The HPE ProLiant DL380 Gen11 NVIDIA L40S Configure-to-Order Server (P60634-B21) is a 2U rack-mount platform engineered to deliver exceptional GPU-accelerated compute density for enterprise AI inference, rendering, and data-intensive workloads. Built on the HPE ProLiant DL380 Gen11 chassis powered by 4th Gen Intel Xeon Scalable processors, the server integrates the NVIDIA L40S GPU — based on NVIDIA's Ada Lovelace architecture — offering 48 GB of GDDR6 ECC memory per GPU, 91.6 TFLOPS of FP32 performance, and dedicated hardware support for Transformer Engine operations critical to large language model inference and training.

The configure-to-order model allows enterprise IT and datacenter architects to tailor the system to specific GPU counts, memory configurations, storage capacities, and network adapters, ensuring alignment with diverse deployment requirements. The DL380 Gen11 platform supports PCIe Gen 5.0 connectivity, DDR5 memory, and HPE's iLO 6 Baseboard Management Controller for advanced remote system management, security, and telemetry. The server accommodates up to 8 NVIDIA L40S GPUs in supported configurations, maximizing throughput per rack unit for demanding GPU compute environments.

Designed for enterprise datacenters across the UAE, GCC, EMEA, and APAC regions, the HPE ProLiant DL380 Gen11 NVIDIA L40S CTO Server is positioned to support AI/ML model development and deployment, professional visualization, virtual workstation hosting, and high-performance simulation tasks. Its compatibility with NVIDIA AI Enterprise software, VMware vSphere, and Red Hat Enterprise Linux virtual GPU frameworks makes it a versatile cornerstone for modern AI-ready infrastructure strategies.

Ideal for

• Large language model (LLM) inference serving and fine-tuning in enterprise AI platforms requiring high GPU memory bandwidth and FP8/FP16 Transformer Engine acceleration
• GPU-accelerated virtual workstation and virtual desktop infrastructure (VDI) delivery, enabling remote professional visualization for design, media, and engineering teams
• High-performance rendering and real-time ray tracing for media production, architecture, and product design workflows leveraging Ada Lovelace RT Cores and Tensor Cores
• Enterprise machine learning training and inference pipelines integrated with NVIDIA AI Enterprise and container-based MLOps frameworks
• Scientific simulation and computational workloads in energy, life sciences, and financial services sectors requiring sustained double-precision and mixed-precision GPU throughput
• Multi-tenant GPU virtualization deployments using NVIDIA vGPU software to share L40S resources across isolated virtual machines in consolidated datacenter environments

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.