GPU cloud provider OpenMetal expanded its hardware catalog with latest-generation Blackwell and H200 accelerators for private server deployments.

OpenMetal has announced the expansion of its v5 hardware catalog with two new private GPU server lines: the RP6000, built on the NVIDIA RTX PRO 6000 Blackwell Server Edition, and the H200, built on the NVIDIA H200 NVL. Both servers run on OpenMetal's Intel Xeon 6000-series and DDR5-6400 v5 platform, bringing dedicated, single-tenant GPU acceleration to AI, machine learning, and high-performance computing workloads. According to Jamie Tischart, CTO of OpenMetal, "With v5 we modernized the foundation of our bare metal and private cloud catalog. Adding the RP6000 and H200 was the natural next step. Customers running AI and HPC workloads get fully dedicated GPUs on the same modern Xeon 6000 platform, with transparent monthly billing and infrastructure they actually control, not throttled, metered slices of someone else's cluster."

Each system pairs dual Intel Xeon 6530P processors with 1TB of DDR5-6400 memory standard, expandable to 2TB, PCIe 5.0 connectivity, NVMe storage options, and 40Gbps private networking. This architecture ensures that the GPU servers' CPU, memory, and I/O subsystems match the rest of the v5 lineup, rather than functioning as a separate, bolted-on accelerator tier.

The RP6000 supports one or two NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs per server. Each GPU provides 96GB of GDDR7 memory, 1.79 TB/s of memory bandwidth, and 24,064 CUDA cores. The platform is designed for inference serving, fine-tuning, rendering, and mixed AI and visualization pipelines that benefit from high VRAM capacity at a favorable cost per GB.

The H200 supports one or two NVIDIA H200 NVL PCIe GPUs per server. Each GPU provides 141GB of HBM3e memory and 4.8 TB/s of memory bandwidth for large-model training and memory-bound inference workloads. Every H200 server includes a five-year NVIDIA AI Enterprise software subscription, giving teams a supported, production-grade stack for deploying AI in their own dedicated environment.

Both GPU server lines are delivered as single-tenant bare metal infrastructure with no shared hypervisor between the customer's workload and the hardware. Customers can run common AI and ML frameworks including PyTorch, TensorFlow, JAX, and Hugging Face Transformers, and can deploy the servers standalone or as part of multi-node GPU clusters over OpenMetal's private network. The RP6000 and H200 are available now in OpenMetal's US East data center in Ashburn, Virginia, with additional locations planned, and proof-of-concept deployments are available for teams that want to validate workloads before committing.