Volume and cluster pricing for larger deployments.
Last reviewed: September 2026
Larger deployments are priced individually. The more detail you can give us about quantity, timeline and workload, the faster we can return an accurate quotation.
01
What to include in your request
A complete request lets us skip a round of clarifying questions and quote directly.
Quantity and target configuration, or the workload you need to run.
Required delivery timeline and deployment location.
Any standards, compliance or procurement requirements.
02
How we scope a deployment
For clusters we design against the workload rather than quoting a parts list, because interconnect and power usually decide real throughput.
GPU selection and node count sized to your training or inference target.
Ethernet or InfiniBand fabric design for the chosen topology.
Power, cooling and rack space validated against your facility.
03
Volume pricing
Pricing improves with quantity, commitment and flexibility on schedule.
Tiered pricing applies above defined quantity thresholds.
Scheduled multi-phase rollouts can be priced as a single commitment.
Tolerance for longer lead times often improves allocation and cost.
04
Timeline
Cluster quotations take longer than single-system quotations because they are engineered rather than configured.
Initial response within one business day.
A scoping call to confirm workload, power and site constraints.
A written proposal with configuration, pricing and delivery schedule.
05
Beyond the hardware
Volume orders usually include services that need to be scoped alongside the equipment.
Rack integration, cabling and on-site bring-up.
Software stack, orchestration and GPU scheduling setup.
Ongoing support, spares strategy and expansion planning.
This page describes our general practice. The terms that apply to your order are the ones stated in your quotation and order documents.
Still need help?
Our team can confirm the exact terms for your configuration, quantity and destination.