TK Lab Blog

LAB ANALYTICS / TRACEABLE ONLY

From rankings to reasons.

An analysis appears only when its sample conditions are satisfied. Incomplete sections show coverage instead of estimated values.

Fan-swap uplift9coolers with 2+ AIDA runs
Rank divergence36records with both metrics
Bench / S5M pairs2comparable configurations
Noise-normalized0requires 3 controlled dBA records
01 / FAN-SWAP UPLIFT

How much a fan swap changes one heatsink

Difference between the current best and lowest AIDA64 result on each heatsink; not a cross-heatsink fan ranking.

02 / RANK DIVERGENCE

R23 and AIDA64 do not produce the same ranking

A large rank gap means short benchmark performance and sustained heat dissipation judge the configuration differently.

03 / IN-CASE DELTA

Open bench → S5M heat-soak loss

Eligible

Only identical cooler, fan configuration, and CPU settings are paired. An environment tag on a noise video does not count as a performance run.

04 / NOISE NORMALIZED

Noise-normalized performance

Awaiting controlled dBA

Distance, noise floor, side panel, and heat soak must match. Video without a dBA value remains an acoustic record but does not enter the ranking.