Skip Navigation
Hardware @lemmy.world
MHLoppy @fedia.io

NVIDIA GeForce RTX 5060 Ti PCI-Express x8 scaling

www.techpowerup.com

NVIDIA GeForce RTX 5060 Ti PCI-Express x8 Scaling

NVIDIA today launched the GeForce RTX 5060 Ti, the latest entry to its performance-segment lineup. You can read all about the card and how it performs in our main review of the card. This card is a direct successor to the RTX 4060 Ti and the RTX 3060 Ti before it. The RTX 3060 Ti and its sibling, the RTX 3060, came with PCI-Express 4.0 x16 host interfaces, giving them the same host bandwidth as the flagship RTX 3090 graphics card from that generation. Then, with its RTX 40-series, NVIDIA changed its mid-range GPUs such that the AD106 silicon driving the RTX 4060 Ti comes with a PCI-Express 4.0 x8 host interface.

When NVIDIA chose to give the RTX 4060 Ti and the AD106 silicon driving it a PCI-Express 4.0 x8 interface, it justified the move by pointing out PCIe Gen 4 being prevalent with desktop platforms going all the way back to the 11th Gen Core Rocket Lake and Ryzen 3000 Zen 2—platforms from 2018 and 2019. With the new RTX 50-series Blackwell generation, NVIDIA is implementing PCI-Express Gen 5, which doubles bandwidth over Gen 4, and chose to give the RTX 5060 Ti and the GB206 silicon driving it a PCI-Express 5.0 x8 host interface.

2 comments

  • We are thrilled to report that running the GeForce RTX 5060 Ti with PCI-Express 4.0 x8 has an insignificant impact on performance. Averaged across all game tests, we see a 2% drop in FPS at 1080p, and an even smaller 1% drop at 1440p. While not exactly relevant to this GPU, even 4K Ultra HD posts only a 1% drop in performance with PCIe Gen 4 x8.

    There is basically no difference with using PCIe 4.0. 1-2% is likely unnoticeable in real life. That's what I've been hearing/reading in the last few years; the PCIe isn't really a bottleneck for dGPUs for consumers.

    Even with PCIe 3.0, techpowerup are seeing a mere ~4% decrease in performance.

    • Yeah this is not a surprising result given their testing on flagship GPUs showed trivial differences too, but it's not bad to verify that reality matches expectations