NVIDIA DGX Rubin NVL8
- 8x NVIDIA Rubin GPUs
- 2x Intel® Xeon® 6776P processors
- 2.3 TB GPU Memory
- NVFP4 Inference: 400 PF
- NVFP4 Training: 280 PF
- FP8/FP6 Training: 140 PF
- NVLink 28.8 TB/s total bandwidth
About
The Foundation for Intelligence at Scale
Agentic AI and reasoning models are redefining the limits of computation. But to take full advantage of these breakthroughs, enterprises need infrastructure that can reason internally and quickly pass information between models and agents. Infrastructure must cost-effectively address communication, coordination, memory movement, and latency across training, post-training, and inference.
NVIDIA DGX™ Rubin NVL8 provides a blueprint for success in the age of agentic AI. Built on the NVIDIA Rubin architecture, DGX Rubin NVL8 is a turnkey infrastructure solution purpose-built to accelerate any AI workload and deliver intelligence at scale.
Specification
Specifications
GPU
8x NVIDIA Rubin GPUs
Total GPU Memory | Bandwidth
2.3 TB | 160 TB/s
Performance
NVFP4 Inference: 400 PF
NVFP4 Training: 280 PF
FP8/FP6 Training: 140 PF
CPU
2x Intel® Xeon® 6776P processors
NVIDIA NVLink Switch System
4x
NVIDIA NVLink Bandwidth
28.8 TB/s total bandwidth
System Power Usage
~24 kW
Networking
8x OSFP ports serving 8x single-port NVIDIA® ConnectX®-9 VPI up to 800 Gb/s NVIDIA Infiniband and Ethernet
2x 400G QSP112 NVIDIA BlueField®-4 DPUs up to 800 Gb/s NVIDIA InfiniBand and Ethernet
Software
NVIDIA DGX OS, Ubuntu, Red Hat Enterprise Linux, Rocky
You May Also Like
Related products
-

NVIDIA HGX A100 (8-GPU)
SKU: N/AMore Information- 8X NVIDIA A100 GPUS WITH 320 GB TOTAL GPU MEMORY
- 6X NVIDIA NVSWITCHES
- 320 GB MEMORY
- 4.8 TB/s TOTAL AGGREGATE BANDWIDTH
-

NVIDIA DGX Station GB300
SKU: N/AMore Information- NVIDIA Grace Blackwell Ultra Desktop Superchip
- Fifth-Generation Tensor Cores
- NVIDIA ConnectX-8 SuperNIC
- NVIDIA DGX OS
- NVIDIA NVLink-C2C Interconnect
- Large Coherent Memory for AI
-

NVIDIA DGX GH200
SKU: DGX GH200-1More Information- 32x NVIDIA Grace Hopper Superchips, interconnected with NVIDIA NVLink
- Massive, shared GPU memory space of 19.5TB
- 900 gigabytes per second (GB/s) GPU-to-GPU bandwidth
- 128 petaFLOPS of FP8 AI performance
Our Customers






























Previous
Next
