1. Home
  2. Platform
CHOOSING

It comes down to who runs the stack.

TABLE 1 — THE THREE TIERS COMPARED
Managed APIPrivate endpointsDedicated clusters
Who runs the stackNachster AINachster AIYou
Hardware sharingShared capacitySingle tenantSingle tenant, bare metal
BillingPer tokenPer reserved GPUCommitted capacity
MinimumOne API key1 node (8 GPUs)32 nodes (256 GPUs)
Lead timeQ2 2027~3–4 months~6 months standard
Best forEvaluation, internal tools, variable inferenceSteady inference, serving your own weightsTraining, sustained large-scale inference

If you need dedicated hardware below 32 nodes, private endpoints are offered per node. Talk to us.

HARDWARE

The NVIDIA B300 standard node.

Every tier runs on the same standard node: eight B300s in a 10U air-cooled chassis, drawing roughly 15 kW per node as a reference figure.

Nachster AI B300 — standard node and first cluster

PlatformHGX B300 NVL16 · 8-GPU node · 10U air-cooled
GPU8 × NVIDIA B300 · 288 GB HBM3e each · ~2.3 TB / node
CPU · memory2 × Intel Xeon 6 · 3 TB DDR5-6400
Local storage2 × 960 GB M.2 (OS) · 8 × 3.84 TB NVMe
Networking8 × ConnectX-8 800G (integrated) · BlueField-3 DPU
Cluster · fabric32 nodes · 256 GPUs · Quantum-X800 InfiniBand · rail-optimized · 1:1
Shared storageDedicated high-performance storage
Power~15 kW per node (reference)
LocationJapan · dedicated, single tenant
Operations · SLAManaged 24×7 · ≥99.5% per allocation unit / month
ContractMulti-year committed capacity · 5-year standard

Standard node configuration; the final BOM is attached to the allocation offer. Later generations — GB300 · Vera Rubin — are allocated on subsequent windows.

CAPACITY

Four cluster sizes.

Every size is built on the same 1:1 non-blocking rail-optimised fabric. GPU memory and load are derived from the per-node reference figures; confirmed values appear in the allocation offer.

TABLE 2 — THE CAPACITY LADDER
SizeGPUTotal GPU memoryIT load (approx.)
32 nodes256~74 TB~0.5 MW
64 nodes512~147 TB~1.0 MW
128 nodes1,024~295 TB~1.9 MW
256 nodes2,048~590 TB~3.8 MW

The first delivery window (Q2 2027) is 32 nodes / 256 GPUs. Larger sizes are allocated on subsequent windows.

Generations and windows.

IN SERVICE

NVIDIA B300

HGX B300 NVL16, allocated by the 8-GPU node. Air-cooled, closed aisle.

FORWARD

NVIDIA GB300

Allocated by the rack, direct liquid cooling. Once a qualified site is in place.

FORWARD

NVIDIA Vera Rubin

Allocated by the rack, direct liquid cooling.

ON REQUEST

Custom cluster

Built to your design. Bring it to the technical meeting.

The allocation unit changes with the generation — a node for B300, a rack for the rack-scale generations. All capacity figures on this site are stated in GPUs; per-rack GPU counts and power density are confirmed at contracting.

NEXT STEP

Tell us what you need. We reply within 48 hours.

Qualified enquiries receive proposed times for a technical meeting within 48 hours, and an allocation offer after that meeting. Information is handled under NDA.