DGX Spark ×2
NVIDIA's Grace Blackwell desktop system for running, fine-tuning and serving large models.
256 GB across two units: the ~405B class at FP4.
Exclusive bare metal. Your hardware is never shared, and you get full root access.
Hardware, one-time
$13,000
Management & rack
2 units × $99 = $198 per month
How paying works
Place a pre-order. We confirm availability and send an invoice. Nothing is charged online.
Racked and managed by us in Asia or the US. Prices in USD, excluding taxes and import duties.
How it's connected
Two units linked by one direct ConnectX-7 cable (200 GbE). No switch needed.
Cabled and validated by us before handover. Switched configurations for larger clusters are available on request.
Total unified memory: 256 GB
Specifications (per unit)
- NVIDIA GB10 Grace Blackwell Superchip
- 20-core Arm CPU (10× Cortex-X925 + 10× Cortex-A725)
- 128 GB LPDDR5x unified memory · 273 GB/s
- Up to 1 PFLOP FP4 AI compute (with sparsity)
- 4 TB NVMe storage, self-encrypting
- ConnectX-7 · 2× QSFP ports, up to 200 Gb/s each
- DGX OS (Ubuntu-based) · ARM64
What you can run
The largest model classes that fit in memory, by configuration. Larger models are split across units over ConnectX-7, so speed depends on the interconnect and software.
Approximate, depends on quantization and context length.
| Configuration | Unified memory | FP8 (~1 byte per parameter) | FP4 (~0.5 byte per parameter) | Source |
|---|---|---|---|---|
| 1× DGX Spark | 128 GB | ~70B comfortably | Up to ~200B | NVIDIA |
| 2× DGX Spark This configuration | 256 GB | ~200B | ~405B class (e.g. Llama 3.1 405B) | NVIDIA |
| 3× DGX Spark | 384 GB | ~300B | ~600B class | Our estimate |
| 4× DGX Spark | 512 GB | ~400B | Up to ~700B (e.g. 671B MoE models) | NVIDIA |
What else it's good for
Fine-tuning and LoRA
NVIDIA rates one unit for fine-tuning models up to 70B with memory-saving methods such as LoRA and QLoRA. Smaller models can be fully fine-tuned.
Multi-agent development
Run several models and agents side by side on hardware you control, with no per-token bill and no data leaving your machine.
Local inference serving
Serve models to your team over a private endpoint with tools such as vLLM, TensorRT-LLM or Ollama.
Racked and managed in Asia or the US
We receive, burn in and rack your machine, then manage it: monitoring, reboots, reimaging and remote hands. You keep full root access and can take delivery whenever you want.
How managed hosting worksVolume pricing for larger configurations
Need more than four Sparks, several dedicated CI machines, or a mix? Tell us the configuration and we'll quote it.
DGX Spark ×2
Place a pre-order. We confirm availability and send an invoice. Nothing is charged online.