GPU comparison
B300 vs B200
B300 (Blackwell Ultra) is the better inference GPU: 288 GB instead of 180 GB and 1.5× the dense FP4 throughput, for roughly 10% more per month. B200 keeps strong FP64 and the same 4.5 PF of FP8, so it stays the sensible choice for classic training and mixed HPC work.
- Updated Nov 4, 2026
- Specs: vendor datasheets, dense Tensor figures
- Prices: live CoreGPU catalog
| Spec | NVIDIA HGX B300 | NVIDIA HGX B200 |
|---|---|---|
| Architecture | Blackwell Ultra | Blackwell |
| GPU memory | 288 GB HBM3e | 180 GB HBM3e |
| Memory bandwidth | 8 TB/s | 8 TB/s |
| FP8 dense | 4.5 PF | 4.5 PF |
| FP4 dense | 13.5 PF | 9 PF |
| Interconnect | NVLink 5 · 1.8 TB/s | NVLink 5 · 1.8 TB/s |
| Power (TDP) | 1,100 W | 1,000 W |
| Price per GPU / month | $4,899 | $4,449 |
| Price per GB of memory | $17.01 | $24.72 |
| Price per FP8 PFLOPS | $1,089 | $989 |
| Availability | Only 9 left | Back Nov 20 |
Green marks the better value on each line. Prices are per GPU per month at CoreGPU, at least 25% below the market median.
- FP4/FP8 inference of reasoning and agent models
- Models or KV caches that need more than 180 GB per GPU
- Fewer GPUs for the same model thanks to 288 GB
$4,899 /GPU/moConfigure B300
- Pre-training and fine-tuning where FP8 is the target precision
- Codes that need double precision (FP64)
- The lower price when 180 GB is enough
$4,449 /GPU/moJoin waitlist B200