GPU rental · request capacity
NVIDIA's deskside AI computer: one Grace Blackwell Ultra superchip with 748 GB of coherent memory, 252 GB of it fast HBM3e on the GPU. NVIDIA designs it, and Dell, HP, ASUS, MSI, GIGABYTE and Supermicro build and sell it.
Specifications
Good for: Running models up to about a trillion parameters at 4-bit on one machine, serving a team from one box, and fine-tuning models too big for any single card.
| Machine | NVIDIA DGX Station (GB300) |
|---|---|
| Memory | 252 GB HBM3e on the GPU + 496 GB LPDDR5X on the CPU (748 GB coherent) |
| GPU | 1× Blackwell Ultra, 7.1 TB/s memory bandwidth |
| CPU | Grace, 72 Arm Neoverse V2 cores, 396 GB/s memory bandwidth |
| CPU to GPU | NVLink-C2C, 900 GB/s |
| AI compute | Up to 20 PFLOPS FP4 (15 without sparsity) |
| Networking | ConnectX-8 SuperNIC, 2× 400 Gb/s |
| GPU partitions | Up to 7 (MIG) |
| Power | 1,600 W, needs a 20 A circuit |
| Form | Liquid-cooled deskside tower |
| Software | Ubuntu 24.04 with NVIDIA's AI tools (DGX OS) |
| Architecture | Grace Blackwell Ultra |
| Released | 2026-03 |
| Price to buy | From about €89,000 (base builds, 2026-10) |
| Availability | Request capacity: register your requirement in the console |
| Tenancy | Dedicated machine, your traffic only, full SSH access |
| Pricing | Quoted per deployment |
Public specifications from the maker. See every GPU we offer.
Community data
| Test | DGX Station | Compared with | Source |
|---|---|---|---|
| gpt-oss-20b, 1 request | 530 to 536 t/s | RTX PRO 6000: 241 to 244 t/s · DGX Spark: 50 t/s | StorageReview, vLLM |
| Qwen3 Coder 30B A3B (FP8), 1 request | 310 to 319 t/s | RTX PRO 6000: 169 to 208 t/s · DGX Spark: 55 t/s | StorageReview, vLLM |
| gpt-oss-120b, 1 request | about 380 t/s | H200 NVL: about 200 t/s · RTX PRO 6000: about 170 t/s · DGX Spark: about 35 t/s | StorageReview, vLLM (read from the chart) |
| gpt-oss-120b, 128 requests at once | about 9,300 t/s in total | H200 NVL: 3,376 t/s · RTX PRO 6000: 3,064 to 3,384 t/s · DGX Spark: 500 t/s | StorageReview, vLLM |
| Llama 2 7B Q4_0, llama.cpp (our decode ladder) | about 300 t/s (estimate) | B200: 298 t/s · GH200: 312 t/s · RTX 5090: 290 t/s | llama.cpp scoreboard, no DGX Station entry yet |
StorageReview ran vLLM with 512 tokens in and 512 out, one run each, on the MSI and ASUS builds. Not our measurements. Results vary with the engine, the quantization and the settings.
Who builds it
| Maker | Machine |
|---|---|
| ASUS | ASUS ExpertCenter Pro ET900N G3 |
| Dell | Dell Pro Max with GB300 |
| GIGABYTE | GIGABYTE W775-V10 |
| HP | HP ZGX Fury AI Station |
| MSI | MSI XpertStation WS300 |
| Supermicro | Supermicro Super AI Station (GB300) |
Models
Estimated from each model's own files and config (252 GB on the GPU): language models · image models · video models · embedding models.
How it works
That is how the DGX Spark offer was built. Ask for this GPU, or Talk to an engineer first.
Available today
AxForge already runs Blackwell: the GB10 Grace Blackwell superchip inside the NVIDIA DGX Spark, available now in Málaga, Spain (eu-es-1) by the hour, week, month or year from €0.85/hour, with community benchmark data on the GB10 page.
FAQ
You can ask for one in the AxForge Console: Ask for this GPU. Registered demand decides which machines we add next. Dedicated DGX Spark (GB10) capacity is available now from €0.85/hour.
Yes, as dedicated DGX Station capacity asked for through the AxForge Console and sized to your workload, on hardware AxForge runs in Europe. Registered demand decides which machines we add next.
GB10 Grace Blackwell, in the NVIDIA DGX Spark, available now in Málaga, Spain (eu-es-1) from €0.85/hour, with community benchmark data on the GB10 page.
DGX Station capacity is quoted per deployment. Tell us the model sizes, traffic and term, and we scope the machine and the price with you in EUR before anything is billed.
Sign in, choose the GPU and describe what you need. Registered demand decides which machines we add next, and the people who asked are the first to be offered them.
NVIDIA DGX Spark (GB10): dedicated, 128 GB unified memory, full SSH access, from €0.85/hour by the hour, week, month or year. Other GPUs can be asked for through the console.
Yes. AxForge runs its own hardware in European regions: eu-se-1 · Stockholm for serverless inference and RTX 3060 rentals, eu-es-1 · Málaga for dedicated DGX Spark workloads.
Explore