Bare-Metal GPU Servers

8x NVIDIA L20 — 384 GB for Cost-Efficient LLM Inference

Eight NVIDIA L20 GPUs on bare metal: 48 GB of GDDR6 per card at 864 GB/s and Ada Lovelace Tensor Cores with FP8, on a dual Xeon Gold 6226R host with 384 GB RAM. No hypervisor. No setup fees.

See Pricing & Order

From €149/day (8 GPUs) · 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 Inference at Scale

48 GB per GPU and the memory bandwidth of the L40S, at a lower price per GPU

LLM Inference & Serving

Serve 70B-class models across eight GPUs with vLLM, SGLang or TensorRT-LLM, or one 7B–13B model per GPU for high-throughput endpoints.

Fine-Tuning

LoRA and QLoRA fine-tunes with PyTorch, Hugging Face PEFT and DeepSpeed on fourth-generation Tensor Cores.

RAG & Embeddings

Run embedding models, rerankers and generators side by side, each on its own GPU, with memory to spare for batches and caches.

Computer Vision

Detection, segmentation and OCR pipelines at high batch sizes, deployed with TensorRT and Triton Inference Server.

Generative Media

Stable Diffusion and Flux image generation and speech models in ComfyUI or Hugging Face Diffusers.

Lower Cost per GPU

Memory-bound LLM decoding depends mostly on memory bandwidth, where the L20 matches the L40S at 864 GB/s.

Software Compatibility

Verified compatibility with the standard inference and AI stack

vLLM
LLM Serving
SGLang
LLM Serving
TensorRT-LLM
Inference Optimization
Triton
Inference Server
Ollama
Local LLMs
PyTorch
Deep Learning
TensorFlow
Machine Learning
Transformers
Hugging Face
PEFT
Fine-Tuning
DeepSpeed
Training Optimization
CUDA 12
Toolkit & cuDNN
ONNX Runtime
Inference
RAPIDS
Data Science
ComfyUI
Generative AI
Diffusers
Generative AI

NVIDIA L20 Server Pricing

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

Configuration GPU Memory RAM Storage Daily Weekly Monthly Per GPU/mo Status
8x L20 2x Xeon Gold 6226R 384 GB 384 GB 2000 GB NVMe €149 €799 €2,900 €363 Order
CPU
2x Intel® Xeon® Gold 6226R Processor 16C/32 3.9 GHz
RAM
384 GB
Storage
2000 GB NVME
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 1 Business Day

Your server is provisioned within 1 business day of your order, with no setup fees and no hypervisor.

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 is provisioned within 1 business day of placing your order. A configuration that is busy shows the date it is available from in the pricing table.
How much GPU memory does the L20 have?
Each NVIDIA L20 carries 48 GB of GDDR6 with 864 GB/s of bandwidth, so the server has 384 GB of GPU memory in total. That serves a 70B-parameter model in FP16 across the eight GPUs, or a separate 7B–13B model on each card.
How does the L20 compare with the L40S?
Both are Ada Lovelace GPUs with 48 GB of GDDR6 and 864 GB/s of memory bandwidth. The L20 has about two thirds of the L40S's CUDA cores (11,776 against 18,176), so compute-bound training is slower, while memory-bound LLM decoding depends mostly on memory bandwidth, where the two cards are equal. The L20 costs less per GPU.
Is there NVLink between the GPUs?
No. The L20 does not support NVLink, so the eight GPUs communicate over PCIe. Tensor parallelism in vLLM and data-parallel training with PyTorch DDP or DeepSpeed work over PCIe.
Which CPU, RAM and storage ship with the server?
Two Intel Xeon Gold 6226R processors (16 cores / 32 threads each), 384 GB of system RAM and 2000 GB of NVMe storage.
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?
L20 servers ship with Ubuntu 22.04 LTS or Ubuntu 24.04 LTS. 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 Serving Models on NVIDIA L20

384 GB of GPU memory on bare metal. No setup fees.

See Pricing & Order View All GPU Servers