Posts tagged: IT Infrastructure
-
DGX Spark 64GB: Specs, Price, and the Cable to Pair Two
NVIDIA announced the DGX Spark 64GB on October 2, 2026: the same GB10 Grace Blackwell Superchip, the same ConnectX-7 networking, the same DGX OS and software stack as the...
-
NJAAKK-N911 vs NJAAKK0006: DGX Spark Cable Part Numbers
If you have searched for a DGX Spark cable you have run into a wall of part numbers: NJAAKK-N911 on the retail listings, NJAAKK0006 and LMTQF022-SD-R in NVIDIA's documentation,...
-
DGX Spark vs Strix Halo vs Gorgon Halo: Local LLM Benchmarks
DGX Spark vs Strix Halo vs Gorgon Halo, tested with the same llama.cpp build: token speed, prompt processing, bandwidth, clustering, price and which to buy.
-
DGX Spark Shutdown Under Load: Our 8-Unit GB10 Diagnosis
Why our DGX Spark and GB10 units powered off at the start of long prompts, the nvidia-smi clock lock that stopped it, and a reboot fix for memory bandwidth.
-
DGX Spark Firmware Update: What We Learned on 8 GB10 Units
Our DGX Spark firmware update on 8 GB10 units: mixed firmware, a unit that shut itself off, an 11% burn drop, and faster LLM serving.
-
What Cable Do I Need to Connect Two DGX Sparks?
One 0.5 m QSFP112 passive DAC connects two DGX Sparks, Port 0 to Port 0. Which cable and why, which GB10 boxes take it, and the bring-up checks that prove 200G.
-
DGX Spark Cluster: 4 to 8 Nodes, Switch vs Ring
Four DGX Sparks need a switch, three can ring. How we wired eight GB10s on two MikroTik switches: parts list, NCCL settings, and what eight nodes bought us.
-
3 DGX Spark Ring: Port-by-Port Wiring Guide
Wire three DGX Sparks in a switchless ring: which QSFP port goes where, why Port 0 to Port 0 breaks NVIDIA's netplan, and the checks that prove 200G.
-
From Two DGX Sparks to a Switched Fabric: Cluster Topology
How to go from two DGX Sparks to a switched four-node GB10 cluster: the CRS812 breakout trap, 196 Gb/s RoCE measured, four-node vLLM results, and a free guide.
-
tmux vs cmux vs herdr: 2026 Terminal Multiplexer Comparison
tmux vs cmux vs herdr for AI coding agent workflows: persistence, SSH, agent state, automation APIs, telemetry and security posture compared.
-
Omarchy Quattro vs NixOS: 2026 Developer Comparison
Omarchy 4.0 Quattro vs NixOS 26.05 for developers: install, declarative config, rollback, NVIDIA, AI coding agents and security posture compared.
-
Mac Studio M5 Ultra vs DGX Spark: 1.2 TB/s vs 273 GB/s
The M5 Ultra brings 1.2 TB/s of bandwidth against the DGX Spark 273 GB/s, but CUDA decides most buys. Compare both for local LLM work and get the FIPS facts.
