News — ConnectX-7

The MMA4Z00-NS400 is the NIC-side module, not the switch-side module. Which ConnectX-7, ConnectX-8 and BlueField-3 ports take it, why it does not go in a Quantum-2 switch or a DGX H100 front cage, and the fiber you need with it.

A compatible MMA4Z00-NS400 or MMA4Z00-NS is safe to run on ConnectX-7 and Quantum-2 when it's coded correctly, DOM/DDM tested, and warrantied. Full spec comparison, the NVIDIA/Mellanox part-number transition, and what to verify before you buy.

Genuine NVIDIA OEM 800G optics are running 40+ week lead times in 2026 due to a global EML laser bottleneck — the same bottleneck NVIDIA just spent $4B locking up priority access to. Here's the data, and why compatible MMA4Z00-NS/NS400 modules ship in 7-10 days instead.

NDR InfiniBand, carried by MMA4Z00-NS400 and MMA4Z00-NS optics, is the scale-out fabric behind NVIDIA's own DGX H100/H200 SuperPOD reference architecture. Here's the real problem it solves, who buys it, and how to size an optics order for it.

A straight comparison of US pricing for the NVIDIA MMA4Z00-NS400 flat-top 400G OSFP SR4 module across genuine OEM distributors and compatible sellers, plus the flat-top/finned-top mistake that trips up ConnectX-7 buyers.

The flat-top module goes in the NIC, the finned-top goes in the switch. Here's the rule, the NVIDIA and Mellanox part numbers, and the third variant almost nobody stocks.

DGX Spark has four USB-C ports, an RJ-45 jack, and two QSFP connectors, and only one of them is built for clustering. Here's what each port is actually for, and which cable to use when linking two Sparks.

The NVIDIA DGX Spark is already a remarkable machine on its own — a desktop-sized AI supercomputer packing up to 1 petaFLOP of FP4 AI performance, 128 GB of unified memory, and a full NVIDIA software stack. But there are very real reasons why researchers, developers, and enterprises are looking to link two or more of them together. This post breaks down the why, the hardware limitations, and exactly how far you can scale a multi-Spark cluster. Short answer: the cable and what it costs Two DGX Sparks connect with one 0.5 m passive DAC — $84 or $187 at...

If you picked up an ASUS Ascent GX10, a Dell Pro Max GB10, an HP ZGX Nano AI Station, an MSI EdgeXpert MS-C931, a Lenovo ThinkStation PGX, a Gigabyte AI TOP ATOM, or an Acer Veriton GN100 instead of an NVIDIA-branded DGX Spark, you're probably searching for the same answer: does the DGX Spark cable work with your box? Short answer: yes. All eight of these systems are built on the same NVIDIA GB10 Grace Blackwell Superchip reference design as the DGX Spark, with an identical ConnectX-7 200GbE port on the back (sometimes called the QSFP/CX7 port, sometimes just labeled...

The NJAAKK-N911 (Amphenol) and LMTQF022-SD-R (Luxshare) are two manufacturer versions of the same NVIDIA-validated 400G QSFP112 passive DAC stacking cable for DGX Spark. Both connect the dedicated QSFP112 stacking port and deliver the same 400G performance. The only difference is the manufacturer and wire gauge (32AWG vs 30AWG). This guide explains when to use each.