Actually Hardcore Overclocking (Buildzoid)

Taking Power Rail Resistance Measurements of an RTX 5080

The video walks through a hands‑on investigation of an Inno3D RTX 5080, a stripped‑down, budget‑oriented variant that cannot adjust its power limit. Buildzoid bought the card on sale to compare its performance against his aging RX 970, only to discover the newer GPU delivers virtually the same frame rates, especially in 1 % low‑FPS scenarios, making the price‑to‑performance ratio questionable. Using the display‑port housing as a reliable ground reference, he probes the card’s power delivery network. The Vcore rail shows roughly 0.26 Ω resistance, while the MSVDD rails sit near 0.85 Ω each, and combined they read about 1.1 Ω. Memory phases exhibit about 7.8 Ω each, confirming a three‑phase memory subsystem. He also identifies up to ten Vcore power stages, likely driven by an MPS 29816 controller, and maps several input‑filtering capacitors tied directly to the 12‑V high‑current connector. A recurring theme is the inability to modify the power limit. Buildzoid notes, “when the card doesn’t let you change the power limit, you got to wonder what’s wrong with the PCB.” He explores the shunt‑resistor network, discovering zero‑ohm jumpers and a complex layout that makes shorting for a switchable limit non‑trivial. He plans to consult the NCP45492 pinout to devise a mod similar to one he implemented on a 4070 Ti. The findings highlight the RTX 5080’s modest performance gains relative to its cost and the engineering challenges of retrofitting power‑limit controls. For enthusiasts and system integrators, understanding rail resistances and phase architecture informs thermal and stability tuning, while the potential for a DIY power‑limit switch could extend the card’s utility in benchmark rigs.