Price floor Every GPU at least 30% below the market median — re-checked against the marketplace weekly.

See the proof

Compare · prices checked 2026-09-03

B200 vs B300: specs, price per hour, which to rent

NVIDIA B200 (192 GB, $4.204/hr) against NVIDIA B300 (288 GB, $7.875/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.

B200 vs B300 specifications

Public NVIDIA figures (dense, non-sparsity). The last column is B200 relative to B300.

SpecB200B300Difference
ArchitectureBlackwell (2024)Blackwell (2025)
VRAM192 GB HBM3e288 GB HBM3e−33%
Memory bandwidth8,000 GB/s8,000 GB/ssame
FP16 tensor (dense)2,250 TFLOPS2,800 TFLOPS−20%
FP32
CUDA cores
TDP1000 W1100 W−9%
PCIe · NVLinkGen 5.0 · NVLinkGen 5.0 · NVLink
PowerScore (RTX 3090 = 100)15851972−20%
Max GPUs per machine

B200 vs B300 price per hour

Fixed rates from our sheet — every on-demand price is the marketplace median × 0.70, rounded down. Monthly = 730 hours.

RateB200B300Cheaper
On-demand, per GPU-hour$4.204$7.875B200 (−47%)
Interruptible, per GPU-hour$2.102$3.937B200
Reserved (3 mo), per GPU-hour$2.732$5.118B200
On-demand, per month$3,069$5,749B200
Market median (reference)$6.01$11.25
$ per 1,000 FP16 TFLOP-hours$1.87$2.81B200 (better value)
$ per GB of VRAM per hour$0.0219$0.0273B200 (better value)

Try a full month with storage and bandwidth in the GPU cost calculator.

What fits in VRAM: 192 GB vs 288 GB

WorkloadB200B300
Largest LLM in FP16, one card~72B~105B
Largest LLM at 4-bit, one card~235B~405B
Flux dev (FP8, ~17 GB)fitsfits
Wan 2.x 14B video (offloaded)yesyes
70B 4-bit LLM on one cardyesyes

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?

The B300 (Blackwell Ultra) raises memory to 288 GB of HBM3e and boosts FP4 inference throughput; the B200 offers 192 GB at a lower hourly rate. Reserve the B300 for the largest inference deployments; the B200 for most Blackwell training.

  • Cheaper per hour: B200 ($4.204 vs $7.875, −47%).
  • More VRAM: B300 (288 GB vs 192 GB).
  • More FP16 throughput: B300 (about 1.2×).
  • Best value per TFLOP-hour: B200.
  • Best value per GB of VRAM: B200.
  • Multi-GPU: B200 with NVLink · B300 with NVLink.
Is the B300 faster than the B200?

On dense FP16 tensor throughput the B300 leads by about 1.2× (2,800 vs 2,250 TFLOPS). Memory bandwidth matters as much for inference: B200 8,000 GB/s vs B300 8,000 GB/s.

Which is cheaper to rent, the B200 or the B300?

The B200: $4.204/hr on-demand versus $7.875/hr — 47% less. Interruptible rates are $2.102 (B200) and $3.937 (B300). Per TFLOP-hour the better value is the B200.

Which has more VRAM and what does that change?

The B300 has 288 GB versus 192 GB. In LLM terms that is roughly a 105B FP16 model (or ~405B in 4-bit) on one card against 72B FP16 (~235B 4-bit). If the model does not fit, speed is irrelevant.

Can I rent both on PowerGPU right now?

Yes — 53 × B200 and 24 × B300 are online as this page renders, deployable in about 30 seconds, billed per second, paid in crypto with no KYC.

Deploy your first GPU in under a minute

Top up in crypto, benchmark us against your current provider. Per-second billing, fixed prices ≥ 30% below market — cancel by just stopping the instance.