Bare-Metal GPU Servers

NVIDIA H200 — 141 GB HBM3e for LLM Inference & HPC

The memory-upgraded Hopper flagship on bare metal. 141 GB of HBM3e at 4.8 TB/s, an FP8 Transformer Engine, and a 64-core AMD EPYC host with 512 GB RAM. No hypervisor. No setup fees. Ready in 15 minutes.

From €149.35/day (1 GPU) · excl. VAT

ISO 27001 & NEN 7510 Certified Data Rooms Tier 3 Compliant Data Rooms 13+ Years on Market
6x FD Gazellen Award Winner
ISO 27001 & NEN 7510 Certified Data Rooms
Tier 3 Compliant Data Rooms
13+ Years on Market

Built for Large-Scale AI & High-Performance Computing

141 GB of HBM3e at 4.8 TB/s removes the memory wall that limits inference throughput and simulation size

LLM Inference & Serving

Serve 70B-class models in 8-bit on a single GPU, with enough headroom left for a long-context KV cache instead of spilling to host memory — ready for vLLM, SGLang, TensorRT-LLM and Ollama.

Fine-Tuning & Training

Full fine-tunes and LoRA runs that would otherwise need a multi-GPU node. The Hopper Transformer Engine accelerates FP8 and BF16 training in PyTorch, DeepSpeed and NVIDIA NeMo.

HPC & Scientific Simulation

Molecular dynamics with GROMACS and AMBER, CFD, seismic and weather codes. Memory bandwidth — not FLOPS — is the limit for most HPC kernels, and the H200 delivers 4.8 TB/s of it.

RAG & Data Analytics

Keep embedding models, rerankers and vector indexes resident in GPU memory alongside the generator. Accelerate dataframe and ETL pipelines with RAPIDS, cuDF and XGBoost.

Computer Vision & Multimodal

Train and serve detection, segmentation and vision-language models at high batch sizes. High-resolution inputs and long video clips stay on the GPU instead of streaming from host RAM.

Generative Media

Run diffusion and video-generation pipelines end to end — text-to-image, text-to-video and audio models keep every stage of the pipeline in GPU memory.

Software Compatibility

Verified compatibility with the standard AI, inference and scientific computing stack

vLLM
LLM Serving
SGLang
LLM Serving
TensorRT-LLM
Inference Optimization
Triton
Inference Server
Ollama
Local LLMs
PyTorch
Deep Learning
TensorFlow
Machine Learning
JAX
Numerical Computing
Transformers
Hugging Face
DeepSpeed
Training Optimization
NVIDIA NeMo
LLM Framework
CUDA 12
Toolkit & cuDNN
RAPIDS
Data Science
GROMACS
Molecular Dynamics
AMBER
Scientific Computing

NVIDIA H200 Server Pricing

No setup fees. No hidden costs. All prices exclude VAT.

Configuration GPU Memory RAM Storage Daily Weekly Monthly Per GPU/mo Status
1x H200 AMD EPYC 9554P 141 GB 512 GB 8000 GB SSD €149.35 €746.75 €2,986.90 €2,987 Order
CPU
AMD EPYC 9554P - 64C/128T
RAM
512 GB
Storage
8000 GB SSD
Network
1 Gbit

Why LeaderGPU

Trusted by researchers and enterprises across Europe

Bare-Metal Performance

No hypervisor, no virtualization overhead. Direct hardware access for maximum GPU throughput.

Ready in 15 Minutes

Automated provisioning gets your server online fast. Start training, fine-tuning, or serving within minutes of ordering.

Tier 3 Compliant Data Rooms

Enterprise-grade data rooms with ISO 27001 and NEN 7510 certification.

Flexible Billing

Daily, weekly, or monthly rental. No setup fees. Scale your commitment to match your project timeline.

13+ Years / 6x FD Gazellen

Operating since 2012. Among only 78 companies to receive the FD Gazellen award in 2024.

Enterprise Connectivity

Peering via AMS-IX and NL-ix. Four redundant internet connections for maximum uptime.

Data Room Infrastructure

Netherlands-based Tier 3 compliant data rooms with enterprise-grade redundancy

  • Tier 3 compliant data rooms supplier
  • ISO 27001 and NEN 7510 data room standards
  • Each server connected to dual-power feeds (UPS)
  • Generators: N+1
  • Stable 24°C cooling, backup units (N+1)
  • Four redundant internet connections with peering via NL-ix and AMS-IX
  • VESDA very early smoke detection + gas fire extinguishing
  • 24/7 security guards, fingerprint scanners, CCTV, intrusion detection

Supported Operating Systems

Ubuntu 22.04 LTS
Ubuntu 24.04 LTS
Custom OS (1 business day)
24/7 Support

Ordering & incidents around the clock. Technical support Mon-Fri, 9:00-18:00 CET.

Frequently Asked Questions

How quickly will my server be provisioned?
Your server will be provisioned and ready to use within 15 minutes of placing your order. Our automated setup process ensures you can start your workload almost immediately.
How much GPU memory does the NVIDIA H200 have?
The NVIDIA H200 carries 141 GB of HBM3e memory with 4.8 TB/s of bandwidth. That is enough to serve a 70B-parameter model in 8-bit on a single GPU with room left for a long-context KV cache, and it keeps memory-bound inference and large HPC datasets fed without splitting the workload across cards.
How does the H200 compare with the H100?
The H200 is the memory-upgraded member of the same Hopper family. Compared with the H100 NVL it offers 1.5 times more memory (141 GB HBM3e vs 94 GB) and 1.2 times higher memory bandwidth, which translates into up to 1.7 times faster performance on memory-bound workloads such as large language model inference. Compute throughput per card is comparable, so the gain is largest where memory capacity or bandwidth is the bottleneck.
Is the server suitable for AI and LLM workloads?
Yes. With 141 GB of HBM3e on a single GPU you can serve and fine-tune large language models that would otherwise need a multi-GPU node. The Hopper Transformer Engine delivers 3,341 TFLOPS of FP8 Tensor Core throughput and accelerates vLLM, TensorRT-LLM, PyTorch and TensorFlow. Bare-metal access means no virtualization overhead.
Which CPU, RAM and storage ship with the H200 server?
The H200 platform pairs the GPU with an AMD EPYC 9554P (64 cores / 128 threads), 512 GB of system RAM and 8000 GB of SSD storage, so data loading and preprocessing keep up with the GPU.
Are there any setup fees?
No. There are zero setup fees for any rental period — daily, weekly, or monthly. You only pay the rental price listed. All prices are exclusive of VAT.
Which operating systems are supported?
H200 servers ship with Ubuntu 22.04 LTS or Ubuntu 24.04 LTS, provisioned within 15 minutes. Custom OS installations are available on request and require 1 business day. All servers come with root access so you have full control over the software stack.
Where are the servers located?
Our servers are hosted in Tier 3 compliant data rooms in the Netherlands, upheld to ISO 27001 and NEN 7510 standards, with peering via AMS-IX and NL-ix for excellent European and global connectivity.
Trusted & Secure

Payment Options

All transactions are processed via PCI-DSS compliant gateways. SEPA bank transfer also available for EU customers.

Start Computing on NVIDIA H200 Today

141 GB of HBM3e on bare metal. No setup fees. Ready in 15 minutes.