AI Servers
NVIDIA HGX H100 8-GPU Server
The gold standard for large-scale AI training.
Product catalogue
Browse our inventory of AI servers, GPU accelerators, CPUs, RAM, and storage. All units sourced from verified channels — new and certified refurbished.
Showing 57 products
AI Servers
The gold standard for large-scale AI training.
AI Servers
Next-generation HBM3e memory for frontier AI workloads.
AI Servers
Purpose-built AI infrastructure for enterprise.
AI Servers
Turnkey AI supercomputer for enterprise deployment.
AI Servers
8-GPU OCP-ready server for hyperscale AI clusters.
AI Servers
Dell's flagship 8-GPU AI training server.
AI Servers
Liquid-cooled HPC and AI node for exascale workloads.
AI Servers
8-GPU server optimised for large language model training.
AI Servers
Versatile 2U server supporting up to 10 GPUs.
AI Servers
8-GPU AMD EPYC-based server for AI and HPC.
AI Servers
High-density 4U server for GPU-accelerated workloads.
AI Servers
192GB HBM3 per GPU — built for LLM inference at scale.
GPUs & Accelerators
Highest-bandwidth H100 for multi-GPU training clusters.
GPUs & Accelerators
High-throughput inference and training in a PCIe form factor.
GPUs & Accelerators
141GB HBM3e — the highest-memory GPU available.
GPUs & Accelerators
Proven AI compute for training and HPC workloads.
GPUs & Accelerators
Ampere architecture in a standard PCIe slot.
GPUs & Accelerators
Versatile GPU for inference, graphics, and AI workloads.
GPUs & Accelerators
Ada Lovelace GPU for professional visualisation and AI.
GPUs & Accelerators
Mainstream AI inference accelerator for enterprise servers.
GPUs & Accelerators
Efficient inference GPU for mainstream AI deployments.
GPUs & Accelerators
192GB HBM3 — unmatched memory for LLM inference.
GPUs & Accelerators
Dual-die CDNA 2 accelerator for HPC and AI.
GPUs & Accelerators
Intel's AI training accelerator with 96GB HBM2e.
GPUs & Accelerators
Professional workstation GPU for AI and visualisation.
CPUs
96-core server processor for memory-intensive AI workloads.
CPUs
64-core Genoa processor with high clock speeds.
CPUs
128-core cloud-optimised processor for dense deployments.
CPUs
64-core Milan — the workhorse of AI server deployments.
CPUs
64-core Rome processor — proven in hyperscale deployments.
CPUs
High-core-count Xeon for enterprise server deployments.
CPUs
60-core Sapphire Rapids with PCIe Gen5 and HBM support.
CPUs
40-core Ice Lake Xeon for balanced compute and memory.
CPUs
144-core Granite Rapids for next-gen AI infrastructure.
CPUs
32-core Sapphire Rapids for cost-effective server builds.
CPUs
16-core high-frequency Genoa for latency-sensitive workloads.
RAM & Memory
High-capacity registered ECC memory for AI servers.
RAM & Memory
128GB RDIMM for maximum memory density per socket.
RAM & Memory
High-speed DDR5 for next-generation server platforms.
RAM & Memory
Ultra-high-capacity LRDIMM for memory-optimised servers.
RAM & Memory
64GB DDR4 RDIMM for existing enterprise server fleets.
RAM & Memory
Reliable enterprise memory for existing server fleets.
RAM & Memory
High-density LRDIMM for memory-intensive legacy platforms.
RAM & Memory
Cost-effective ECC memory for server refresh projects.
RAM & Memory
High-bandwidth memory for GPU and accelerator upgrades.
RAM & Memory
Compact ECC memory for edge AI servers and workstations.
Storage
High-endurance NVMe for AI training data pipelines.
Storage
15.36TB NVMe for high-capacity AI storage nodes.
Storage
Balanced capacity and performance for mixed workloads.
Storage
PCIe Gen5 NVMe for next-generation AI server platforms.
Storage
EDSFF form factor NVMe for high-density storage shelves.
Storage
High-capacity SAS drives for bulk data storage.
Storage
24TB nearline SAS for maximum bulk storage density.
Storage
Enterprise SAS SSD for legacy storage infrastructure.
Storage
Massive storage expansion for AI data lake deployments.
Storage
Turnkey all-flash array for AI training data pipelines.
Storage
Compact M.2 NVMe for workstation and edge AI builds.
We maintain deep inventory beyond what's shown here. Send us your requirements and we'll respond within 48 hours.
Submit a Requirements List