Dell Pro Max with GB10 Cluster Cable: 0.5m QSFP112 400G DAC
To cluster two Dell Pro Max workstations with the GB10 Grace Blackwell superchip you need one 0.5m QSFP112 400G passive DAC cable in the QSFP112 port on each unit. Dell's GB10 system carries the same NVIDIA ConnectX-7 NIC as the DGX Spark, so it takes the exact cable NVIDIA approves in its Spark Stacking documentation (Amphenol NJAAKK0006 / Luxshare LMTQF022-SD-R spec). Petronella Technology Group, Inc. sells it at $159 with free US shipping, in stock, shipping in 1 to 3 business days. As of August 2026, retail listings for the same NVIDIA-approved spec run $176 to $285, NVIDIA's own marketplace lists it at $99.99 but is chronically out of stock, and Amphenol's factory-direct store shows its whole QSFP112 line on backorder.
The Dell Pro Max with GB10 brings NVIDIA's GB10 desktop AI platform into Dell's enterprise workstation line, which means IT teams often deploy them in pairs or small clusters under managed procurement. Whichever way you buy the machines, the interconnect is the same single niche part, and it is the one component distributors rarely keep on the shelf.
| Cable | 0.5m QSFP112 400G passive direct-attach copper (DAC) |
| Approved spec | Amphenol NJAAKK0006 / Luxshare LMTQF022-SD-R (NVIDIA Spark Stacking docs) |
| Fits | Both QSFP112 ports on the Dell Pro Max GB10 ConnectX-7 NIC |
| For two units | One cable |
| For a three-node ring | Three cables (switchless, each box uses both ports) |
| Price | $159, free US shipping (in stock, ships in 1 to 3 business days) |
Clustering your Dell Pro Max GB10 units: how many cables and when you need a switch
Each Dell Pro Max GB10 has two QSFP112 ports, so it follows the NVIDIA-documented GB10 topology: two units link directly with one cable, and three units form a switchless ring with three cables (NVIDIA's "Connect Three DGX Spark in a Ring Topology" playbook, each box using both ports for full pairwise connectivity). Three nodes is the switchless ceiling - with only two ports per unit, four or more units need a 200G-class QSFP switch. Clustering pools memory to run a model too large for one 128 GB node; it is a capacity play, not a speed boost, since each link runs around 25 GB/s.