Side-by-side GPU comparison: specs, memory, compute performance, and live cloud pricing.
Verdict
B200 has 64 GB more VRAM, making it better suited for large models and long context windows. For compute-bound workloads like training, B200 delivers 1.2× higher FP16 throughput. B200 supports a broader range of models (659 vs 649 from this catalog), giving more flexibility.
Specifications
| B200 | Gaudi 3 | |
|---|---|---|
| VRAM | 192 GB | 128 GB |
| VRAM Type | HBM3e | HBM2e |
| Memory Bandwidth | 8.0 TB/s | 3.7 TB/s |
| FP16 Performance | 2250 TFLOPS | 1835 TFLOPS |
| Manufacturer | NVIDIA | Intel |
| FP8 Support | Yes | Yes |
| FP4 Support | Yes | No |
Price / Performance
Based on cheapest single-GPU on-demand pricing. Lower $/TFLOP = better compute value; lower $/GB = better memory value.
| B200 | Gaudi 3 | |
|---|---|---|
| $/hr (cheapest) | $5.49 | — |
| $/TFLOP (compute value) | $0.0024 | — |
| $/GB VRAM (memory value) | $0.0286 | — |
Cloud Pricing
Cheapest on-demand price per provider (single GPU).
Model Compatibility
Models from the catalog that fit on each GPU, grouped by required precision.
B200 (659 models)
Gaudi 3 (649 models)
You might also compare…
Pricing data refreshed hourly · Last updated April 11, 2026 · Browse all comparisons