
NVIDIA DGX Station
Deskside AI supercomputer for developers and research teams working on large models.
- Superchip: NVIDIA GB300 Grace Blackwell Ultra Desktop
- Memory: 748 GB coherent memory
- Networking: ConnectX-8 SuperNIC
Compact desktop AI supercomputer for prototyping, fine-tuning and running models locally.
NVIDIA DGX Spark is a compact desktop AI supercomputer built on the GB10 Grace Blackwell Superchip. It gives developers up to 1 petaFLOP of FP4 AI performance and up to 128 GB of unified memory to build and run models and agents locally.
It handles fine-tuning of models up to 70 billion parameters and inference up to 200 billion parameters, ships with the NVIDIA AI software stack, and two to four units can be linked through the built-in ConnectX-7 networking.
Up to 1 PFLOP of FP4 AI performance with fifth-generation Tensor Cores.
64 GB or 128 GB of coherent LPDDR5X shared by CPU and GPU.
200 Gb/s networking to connect up to four DGX Spark systems.
DGX OS with NVIDIA NIM, frameworks, libraries and pre-trained models.
150 × 150 × 50.5 mm, 1.2 kg, 140 W GB10 TDP.
NVIDIA Agent Toolkit for building safe, always-on local agents.
| Superchip | NVIDIA GB10 Grace Blackwell |
|---|---|
| CPU | 20-core Arm (10× Cortex-X925 + 10× Cortex-A725) |
| GPU | Blackwell generation; 5th-gen Tensor Cores, 4th-gen RT Cores |
| Tensor performance | Up to 1 PFLOP FP4 |
| System memory | 64 GB or 128 GB LPDDR5X unified |
|---|---|
| Memory bandwidth | 273 GB/s |
| Storage | Up to 4 TB NVMe M.2 with self-encryption |
| Networking | ConnectX-7 @ 200 Gb/s; 1× RJ-45 10 GbE |
|---|---|
| Wireless | Wi-Fi 7, Bluetooth 5.4 |
| USB | 4× USB Type-C |
| Display | 1× HDMI 2.1a, up to 3× DisplayPort over USB-C |
| Power supply | 240 W |
|---|---|
| GB10 TDP | 140 W |
| Dimensions | 150 × 150 × 50.5 mm |
| Weight | 1.2 kg |
| Noise | 35 dB(A) operating, 19 dB(A) idle |
| Operating system | NVIDIA DGX OS |
Tell us your workload and timeline for the NVIDIA DGX Spark. We reply within one business day with a configuration, pricing and lead time.