Compare · prices checked 2026-09-03
H100 SXM vs B200: specs, price per hour, which to rent
NVIDIA H100 SXM (80 GB, $1.587/hr) against NVIDIA B200 (192 GB, $4.204/hr): public specs side by side, live fixed prices from our sheet, what fits in each card's VRAM, and a verdict written for real workloads — both are rentable right now.
H100 SXM vs B200 specifications
Public NVIDIA figures (dense, non-sparsity). The last column is H100 SXM relative to B200.
| Spec | H100 SXM | B200 | Difference |
|---|---|---|---|
| Architecture | Hopper (2022) | Blackwell (2024) | — |
| VRAM | 80 GB HBM3 | 192 GB HBM3e | −58% |
| Memory bandwidth | 3,350 GB/s | 8,000 GB/s | −58% |
| FP16 tensor (dense) | 990 TFLOPS | 2,250 TFLOPS | −56% |
| FP32 | 67.0 TFLOPS | — | — |
| CUDA cores | 16,896 | — | — |
| TDP | 700 W | 1000 W | −30% |
| PCIe · NVLink | Gen 5.0 · NVLink | Gen 5.0 · NVLink | — |
| PowerScore (RTX 3090 = 100) | 697 | 1585 | −56% |
| Max GPUs per machine | 8× | 8× | — |
H100 SXM vs B200 price per hour
Fixed rates from our sheet — every on-demand price is the marketplace median × 0.70, rounded down. Monthly = 730 hours.
| Rate | H100 SXM | B200 | Cheaper |
|---|---|---|---|
| On-demand, per GPU-hour | $1.587 | $4.204 | H100 SXM (−62%) |
| Interruptible, per GPU-hour | $0.793 | $2.102 | H100 SXM |
| Reserved (3 mo), per GPU-hour | $1.031 | $2.732 | H100 SXM |
| On-demand, per month | $1,159 | $3,069 | H100 SXM |
| Market median (reference) | $2.27 | $6.01 | — |
| $ per 1,000 FP16 TFLOP-hours | $1.60 | $1.87 | H100 SXM (better value) |
| $ per GB of VRAM per hour | $0.0198 | $0.0219 | H100 SXM (better value) |
Try a full month with storage and bandwidth in the GPU cost calculator.
What fits in VRAM: 80 GB vs 192 GB
| Workload | H100 SXM | B200 |
|---|---|---|
| Largest LLM in FP16, one card | ~32B | ~72B |
| Largest LLM at 4-bit, one card | ~123B | ~235B |
| Flux dev (FP8, ~17 GB) | fits | fits |
| Wan 2.x 14B video (offloaded) | yes | yes |
| 70B 4-bit LLM on one card | yes | yes |
Rules of thumb: ~2.4 GB per billion parameters in FP16 all-in, ~0.62 GB in 4-bit. Full tables in the VRAM guide.
Verdict: which should you rent?
Blackwell delivers about 2.3× the dense FP16 throughput of the H100 SXM with 2.4× the memory; the H100 is cheaper per hour and far deeper in supply. B200 for wall-clock time and FP4 serving, H100 SXM for price per FLOP and interruptible availability.
- Cheaper per hour: H100 SXM ($1.587 vs $4.204, −62%).
- More VRAM: B200 (192 GB vs 80 GB).
- More FP16 throughput: B200 (about 2.3×).
- Best value per TFLOP-hour: H100 SXM.
- Best value per GB of VRAM: H100 SXM.
- Multi-GPU: H100 SXM with NVLink · B200 with NVLink.
Related comparisons
- H100 SXM vs A100 SXM4 $1.587 vs $0.583 per hour Compare
- H100 SXM vs H200 $1.587 vs $3.058 per hour Compare
- H200 vs B200 $3.058 vs $4.204 per hour Compare
- H100 SXM vs H100 PCIE $1.587 vs $2.147 per hour Compare
- H100 NVL vs H100 SXM $1.877 vs $1.587 per hour Compare
- B200 vs B300 $4.204 vs $7.875 per hour Compare
H100 SXM vs B200: FAQ
Deeper reading: H100 vs H200 vs B200, RTX 4090 vs RTX 5090, how cloud GPU pricing works.
Is the B200 faster than the H100 SXM?
On dense FP16 tensor throughput the B200 leads by about 2.3× (2,250 vs 990 TFLOPS). Memory bandwidth matters as much for inference: H100 SXM 3,350 GB/s vs B200 8,000 GB/s.
Which is cheaper to rent, the H100 SXM or the B200?
The H100 SXM: $1.587/hr on-demand versus $4.204/hr — 62% less. Interruptible rates are $0.793 (H100 SXM) and $2.102 (B200). Per TFLOP-hour the better value is the H100 SXM.
Which has more VRAM and what does that change?
The B200 has 192 GB versus 80 GB. In LLM terms that is roughly a 72B FP16 model (or ~235B in 4-bit) on one card against 32B FP16 (~123B 4-bit). If the model does not fit, speed is irrelevant.
Can I rent both on PowerGPU right now?
Yes — 69 × H100 SXM and 53 × B200 are online as this page renders, deployable in about 30 seconds, billed per second, paid in crypto with no KYC.