Insights

Perspectives on the infrastructure
behind AI.

Engineering thinking on power, cooling, GPU infrastructure and performance.

Featured editorial

Vol. 04 / Issue 19

Liquid Cooling14 Oct 2026 · 8 min read

Designing liquid cooling for high-density AI racks

Practical considerations for direct-to-chip deployment, coolant distribution units, secondary fluid loops, and facility water integration at scale.

Dr. S. KulkarniPrincipal Thermal Systems

Power & Efficiency
How rack power density shapes electrical architecture
28 Sept 2026 · 6 min read
Grid load dynamics
AI Performance
Beyond installed hardware: what really drives GPU utilisation
19 Sept 2026 · 7 min read
Compute efficiency
AI Performance
Measuring inference: tokens per second, latency and energy per output
04 Sept 2026 · 5 min read
Inference telemetry
Networking
Ethernet or InfiniBand for AI clusters?
22 Aug 2026 · 8 min read
Fabric topologies
GPU Infrastructure
Retrofitting existing facilities for AI workloads
11 Aug 2026 · 6 min read
Brownfield conversion
Standards & Compliance
Commissioning an AI data centre: from systems testing to sustained load
30 Jul 2026 · 9 min read
Commissioning QA
Power & Efficiency
PUE beyond the headline number: what to measure in liquid-cooled halls
14 Jul 2026 · 7 min read
Efficiency metrics
Liquid Cooling
Rear-door heat exchangers or direct-to-chip? Choosing a path to 100kW racks
26 Jun 2026 · 8 min read
Hybrid cooling
Standards & Compliance
Reading ASHRAE thermal guidelines for liquid-cooled hardware
05 Jun 2026 · 6 min read
Thermal guidelines