AxForge
Sign in Get started

GPU rental · request capacity

RTX 5090 rental in Europe: 4-GPU dedicated systems

Systems with four NVIDIA RTX 5090 cards, 32 GB GDDR7 each, as dedicated EU-hosted AxForge machines. Request dedicated capacity for this GPU through AxForge. Registered demand is used to prioritize new capacity.

Available Four cards · 32 GB GDDR7 each Dedicated · quoted per deployment rtx-5090.axforge.ai
Ask for this GPU Available now: DGX Spark (GB10) Register your requirement in the console and AxForge scopes the configuration with you. Dedicated DGX Spark capacity is available now from €0.85/hour.
A machine with an NVIDIA RTX 5090

Specifications

What you get

SystemFour NVIDIA RTX 5090 cards in one dedicated machine
Memory32 GB GDDR7 per card
AvailabilityAvailable: register your requirement in the console
TenancyDedicated machine, your traffic only, full SSH access
PricingQuoted per deployment
RegionEuropean region, placement confirmed with your request

Four cards in one box: run one model per card, or spread a workload across the system. An engineer can help map your models onto the machine when you request capacity.

Community data

The fastest single-stream card on the board

On the same-workload llama.cpp CUDA scoreboard the RTX 5090 decodes at 290.0 t/s with 14,073 t/s prompt processing, the fastest single-stream result on the board, edging even the 96 GB RTX 6000 Pro (274.2 t/s) when the model fits in 32 GB. Roughly 3.8× an RTX 3060, not 10×: the expensive cards earn their keep on capacity and batch throughput, not just one user's tokens.

Same-workload numbers from the llama.cpp CUDA scoreboard (Llama 2 7B Q4_0, tg128, full GPU offload). Community measurements, not ours. Full ladder on the GPU catalogue.

Community data

Modern models at 4k context

Measured by Hardware Corner with llama.cpp at 4k context, one RTX 5090:

ModelQuantDecode @ 4k
Qwen3 30B-A3B MoEQ4226.1 t/s
Qwen3 8BQ4200.4 t/s
Gemma 4 26BQ4180.3 t/s
Qwen3.5 35B MoEMXFP4165.2 t/s
Qwen3 14BQ4123.8 t/s
Qwen3 32B denseQ461.4 t/s
Qwen3.5 27B denseQ458.8 t/s

MoE vs dense, again

226 t/s for a 30B MoE vs 61 t/s for a 32B dense. Same lesson as GB10: active parameters and memory traffic beat headline parameter count.

Software moves the numbers

A newer llama.cpp run of the same 30B MoE reaches 365.7 t/s on a short tg128 profile. It's a different benchmark shape, so we don't mix it into the table, but it shows how far the serving stack moves the result.

Numbers above are single-card, single-stream. Rent one card or a multi-GPU system, as your workload needs. Community measurements, not ours.

Data & privacy

Dedicated, EU-hosted

Like every AxForge machine: your model, your traffic, our hardware. Prompts are never persisted. Only request metadata (token counts, timestamps, status) is kept for billing and operations. Full policy at axforge.ai/privacy.

Models

Which models fit one RTX 5090

Estimated from each model's own files and config (32 GB per card): language models · image models · video models · embedding models.

FAQ

RTX 5090 rental: common questions

Can I rent an RTX 5090 server from AxForge?

Request dedicated capacity for this GPU through the AxForge Console: Ask for this GPU. Registered demand is used to prioritize new capacity, and machines are sized to the workloads that asked for them. Dedicated DGX Spark (GB10) capacity is available now from €0.85/hour.

What is in an AxForge RTX 5090 system?

Four NVIDIA RTX 5090 cards with 32 GB GDDR7 each, rented as one dedicated machine serving only your traffic.

What does RTX 5090 rental cost?

Quoted per deployment for the dedicated 4-GPU system. Register your requirement with model sizes, traffic and term, and AxForge scopes the configuration and price with you in EUR, before anything is billed.

Where will the RTX 5090 systems be hosted?

In an AxForge European region. Dedicated GPU rentals currently run from Málaga, Spain (eu-es-1); placement is confirmed with your request.

Do you keep prompts on a rented RTX 5090 system?

No. Your model, your traffic, our hardware. Prompts are never persisted. Only operational metadata is kept for billing and operations. See the privacy policy.

Need a machine scoped to your workload?

Ask for this GPU Talk to an engineer

Explore

All GPUs

rtx-5090.axforge.ai, the short link for this product → DGX Spark, from €0.85/hour GB10 deep-dive RTX 3060: one card or two, €0.24/hour RTX 6000 Pro: request capacity RTX 3090: request capacity H100: request capacity H200: request capacity B200: request capacity DGX Station: request capacity All GPUs in the catalogue Hosted models → Pricing →
© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms