
Custom Inference Build (from $8,900)
Tell us the model, the latency budget and the traffic curve - we return a costed configuration and hand over a burn-in-tested node. You pay for throughput, not for badge engineering.
MOQ 1 unit
| Type | Build-to-order inference node |
| Base price | From $8,900 (2U, 2x L4 config) |
| GPU options | L4, L40S, RTX 6000 Ada, H100 PCIe |
| CPU options | Xeon Silver/Gold or EPYC 9004 |
| Memory | 128 GB - 1 TB DDR5 ECC |
| Storage | NVMe scratch sized to model cache |
| Networking | 25G standard, 100/400G optional |
| Sizing service | Free: send model + latency/QPS target |
| Validation | Full burn-in + vLLM/TensorRT-LLM smoke test |
| Assembly | Built to order ex-Singapore |
| Warranty | 36 months, SYNDROME AG build |
Tell us the model, the latency budget and the traffic curve - we return a costed configuration and hand over a burn-in-tested node. You pay for throughput, not for badge engineering.
This SKU ships against confirmed allocation — the sourcing desk validates availability per order. Quoted FOB Singapore / Dubai; ask the desk for landed-cost calculation to your port. Each shipment leaves with serial records and export paperwork prepared by our team.
Air and sea freight from Singapore or Dubai under EXW, FOB, CIF or DDP (Incoterms 2020). Export clearance and shipping documents are handled by our hubs. Commercial terms are structured individually for each deal and jurisdiction — your manager will walk you through the options.
New hardware carries a 12-month warranty (refurbished — 6 months). RMA is handled through the nearest hub with advance replacement from stock where available. Serial numbers are recorded for every unit that leaves the warehouse.
