News

Contamination is the leading cause of fiber link failure, and one dirty position on an MPO ferrule presents as a failing module. The inspect-clean-inspect discipline, the right tools, what never to do, and the symptom patterns that should send you to the scope instead of the spares shelf.

Before replacing a transceiver, read its DOM. Transmit and receive power split a dead link into three buckets in under a minute. Per-lane readings, the loss budget arithmetic, and the failures DOM is blind to.

One A40 48GB runs 27–34B models at 23–29 tokens/sec, fits a 70B at Q4 at 12–13.5 tokens/sec, and serves 1,705 tokens/sec of aggregate output on an 8B model at 100 concurrent requests. The measured numbers, the KV cache math behind the 72B cliff, and the three things an Ampere card cannot do.

An NVIDIA A40 48GB works in a Dell PowerEdge R740xd — but the card is not the purchase. The full bill of materials: GPU enablement kit 330-BBMO (US) / 490-BEIX (EMEA), low-profile heat sinks in both CPU sockets, the high-performance fan kit, air shroud and mylar, a dedicated GPU power cable, and PSUs of 1100W or higher — plus the two mistakes that quietly ruin most self-assembled builds.

MTEF-KIT-J is NVIDIA OPN 930-9NRKT-00JN-000, the fixed four-post rack kit for SN3700, SN3700C, SN4600, SN4600C and SN4700. The full MTEF-KIT family crosswalk, why MTEF-KIT-C keeps getting mistaken for it, and the depth-range conflict inside NVIDIA's own manuals.

A field guide to the physical layer in 2026 — SFP through OSFP-XD, plus the PCIe-domain connectors (OCuLink, SlimSAS, MCIO, CopprLink) and the proprietary shells on top of both. Lane counts, speeds, modulation, and what actually plugs into what.

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.

NVIDIA's NVLink 6 switch moves 3.6 TB/s to every GPU and 260 TB/s across a Vera Rubin NVL72 rack. Here is what it actually does, what it makes possible, and why it forces a rebuild of memory, power, cooling and cabling around it.

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.