B300 in Atlanta, Georgia, ready within 4–6 weeks after the deposit: $5.00/GPU·hr on a 24-month term, $4.70 on 36 months, ~19% under Verda's published two-year rate. Written quote within 24 hours. Nothing is owed until you accept it.
Whole nodes only, no fractions, no shared hosts. Terms of 24 or 36 months, take-or-pay, 30% at signing, from a single node. Pick the term and the node count; the written quote names the site and the delivery window.
Dedicated HGX B300 nodes in Atlanta, Georgia, ready for service within 4–6 weeks after the deposit is received. Acceptance benchmarks before billing starts; SLA with service credits per agreement. B300 product page →
Over 20 years of global executive leadership: business management, entrepreneurship, and strategic development across diverse international markets.
Leads a unit that puts datacenter GPU capacity into production. Talks to DCs, engineers, and buyers in one thread — from rack constraints to contract language.
Senior engineer with 10+ years building high-performance systems across autonomous driving, quantitative trading, and large-scale ML inference.
Infrastructure engineer specializing in confidential AI compute, GPU performance engineering, and large-scale inference operations.
Nodes are ready for service within 4–6 weeks after the 30% deposit is received; the exact window is written into your capacity agreement. Tell us your target date in the request and we'll say plainly what's possible. Billing starts only after acceptance benchmarks pass.
Two: 24 months or 36 months, on whole 8-GPU nodes from a single node up, take-or-pay with 30% at signing; the 36-month term carries the lower rate. Month-to-month, quarterly and annual terms are not offered. The rates are on Pricing; the written quote names the site, the delivery window and the exact node configuration.
Requesting a configuration is free and creates no obligation. Your quote — site, delivery window, spec and rate — comes within 24 hours. A 30% prepayment then locks your slot and rate; from that point the reservation is take-or-pay.
Request a configuration → written quote within 24 hours → you accept → KYC, then MSA and order form → 30% prepayment invoice → slot and rate locked, take-or-pay. Remaining 70% per the agreement, monthly invoicing is standard.
A locked reservation is a rental contract, not a hold: the 30% prepayment is credited to your term and is not refunded if you cancel. Until you accept a quote, nothing is owed.
If the delivery window in your agreement is missed, you choose: a rate reduction for the delay, or a full refund of the prepayment if the delay is material. Billing never starts before acceptance, and your rate does not move.
Acceptance criteria are written into your agreement before delivery. If the cluster doesn't pass them, the prepayment is refunded in full. The criteria are spec-based until first delivery: nccl-tests all_reduce inside the node over NVLink and across nodes over the cluster fabric, against the bandwidth thresholds named in the agreement; DCGM diagnostics at level 3; and a burn-in run.
Yes. Nodes in one reservation are delivered together as a single cluster over NVIDIA ConnectX-8 networking; the per-GPU bandwidth is named in your written quote. Reserve two or more and multi-node training works from day one.
Reserved terms: a 30% prepayment invoice after you accept the quote, then monthly invoices per the capacity agreement. Billing starts only after acceptance benchmarks pass. No egress fees.
Service levels are set per cluster in the capacity agreement and backed by service credits for shortfalls. They differ between batches, but every batch carries a production-grade uptime commitment — see Terms for the credit mechanics.
Our nodes are hosted in the United States (Atlanta, Georgia); the hardware is of U.S. origin, so U.S. export controls apply and we follow them: every organization passes screening during quote acceptance, before workloads run.
Written quote in 24h. No payment until you accept.
From the blogHow Much Blackwell Actually Exists?Blackwell shipments set records every quarter — and the median public B300 on-demand price rose 57% since November. We reconstruct how much has actually shipped (~7M packages), how little of it is rentable (two sub-12-month listings), and why prices rise while supply booms.All articles →
NVIDIA B300 product pageRate card and published-rate comparisonAbout TheAI Cloud
Questions before a request: cloud@theai.com
TheAI