Reserve Blackwell capacity.
The NVIDIA B200 is the flagship Blackwell data-center GPU — 192 GB of HBM3e, 8 TB/s of bandwidth, and up to 20 PFLOPS of FP4 compute. Join the waitlist to reserve capacity for large-scale training, fine-tuning, and high-throughput inference.
Reserve capacity now. No commitment — we'll reach out as B200 cards open up in your region.
Free & non-binding. We'll only use your details to contact you about B200 capacity. See our privacy policy.
You tell us how many B200 GPUs you need and your primary use case. As B200 capacity opens up in your region, we contact waitlisters in order, prioritizing larger reservations and earlier signups. There's no payment or commitment to join — you only commit once you confirm a specific cluster offer.
B200 capacity starts at $3.75/GPU-hour for on-demand allocations, with significant discounts for committed multi-month reservations. Pricing is transparent and predictable — no hidden egress fees, no auction-driven price swings. Final pricing depends on contract term, region, and cluster size.
None. Joining the waitlist is free and non-binding. You're under no obligation to purchase when we reach out — we'll simply share availability and pricing for your stated capacity. You can update or withdraw your reservation interest at any time by emailing us.
For small allocations (1–8 GPUs), onboarding typically takes 1–3 business days from offer acceptance. Larger clusters (64+ GPUs) usually onboard in 1–2 weeks depending on networking and storage configuration.