The HGX B200 board with its eight B200 GPUs, the heart of a DGX B200 (shown on its own, outside the chassis). Photo: Pokiiri, CC BY-SA 4.0, via Wikimedia Commons, background removed and resized.DGX B200: 482.3 × 444 mm, 897.1 mm deep. Model-specific illustration, drawn to scale.
NVIDIA · priced at Broadberry
NVIDIA DGX B200
Eight B200 GPUs with 180 GB each, two Xeon Platinum 8570 CPUs, eight 400 Gb/s ConnectX-7 network cards and about 14.3 kW at full load. Priced by Broadberry with three years of support.
Key features
- Eight B200 GPUs with 180 GB each: 1,440 GB of HBM3e
- 18 PFLOPS of dense FP16 by NVIDIA's HGX B200 figures
- Room for a ~576-billion-parameter model at 16-bit, ~2.3 trillion at 4-bit
- Eight 400 Gb/s ConnectX-7 ports and two BlueField-3 DPUs for the cluster network
- 10U, 142 kg, about 14.3 kW at full load, three years of support in the price
Parts and prices
| Part | Qty | Price |
|---|---|---|
| NVIDIA DGX B200 | 1 | $562,154.09 Broadberry, 25 Oct 2026 |
Specifications
NVIDIA DGX B200
- GPUs
- 8 × B200 (180 GB each)
- CPU
- 2 × Intel Xeon Platinum 8570, 112 cores in total
- System memory
- 4 TB DDR5 as configured (NVIDIA: 2 TB, up to 4 TB)
- Storage
- 2 × 1.92 TB M.2 + 8 × 3.84 TB NVMe
- Networking
- 8 × ConnectX-7 400 Gb/s InfiniBand/Ethernet; 2 × BlueField-3
- Form factor
- 10U, 897 mm deep, 142 kg
- Note
- 1,440 GB of GPU memory in total; 3-year support included in the price
- Power
- 14.3 kW (maximum, NVIDIA)
Sources: NVIDIA DGX B200 specifications ↗ · NVIDIA DGX B200 user guide (size) ↗ · NVIDIA HGX B200 (FP16) ↗ · Broadberry configuration ↗
Power adds each part's maximum rated power, before cooling and power losses. Model sizes are memory arithmetic: 2 bytes per parameter at 16-bit, half a byte at 4-bit, keeping a fifth of the memory for the context cache. They say what fits, not how fast it runs.
Can't afford it? Rent it for real.
Cloud providers rent the NVIDIA B200 from $7.99 per GPU an hour, or $2.50 on spot, compared on GlobeGPU (rates of 25 Oct 2026).
Compare NVIDIA B200 prices on GlobeGPU ▸