GoKawiilModder splits DLSS 5 neural rendering across dual RTX 5060 Ti GPUs for big FPS gains
Developer Marcelo Guibout built a ReShade-based tool called MGPU Bridge that offloads Nvidia's DLSS 5 Neural Rendering step to a second GPU, letting one card handle game rendering while the other applies neural post-processing. Demonstrated on Cyberpunk 2077 and The Blood of Dawnwalker using two RTX 5060 Ti cards, the setup reportedly boosted neural-rendered frame rates by up to 127%, though it requires a second monitor and adds display latency. The project is available on GitHub for others to test.
GoKawiil's interpretation of the reporting above, not reported fact.
This mirrors the old dedicated PhysX GPU era, showing enthusiasts are already finding creative workarounds to offset the performance cost of Nvidia's new neural rendering pipeline. It hints at possible future multi-GPU configurations optimized for AI-driven graphics workloads, even though the added latency and hardware requirements make it impractical for most users today.
- A modder split DLSS 5's neural rendering workload onto a second GPU via a custom ReShade tool.
- Tests on dual RTX 5060 Ti cards showed up to 127% higher neural-rendered frame rates.
- The technique needs two monitors and introduces extra display latency, limiting practical use for now.
MSI GeForce RTX 5060 Ti 16GB — The mod's demo runs on a pair of RTX 5060 Ti 16GB cards, one rendering the game and the other handling DLSS 5 neural post-processing. With 16GB of VRAM it's a sensible single-card pick too, and a second one opens the door to experimenting with this dual-GPU offload trick yourself.
See MSI GeForce RTX 5060 Ti 16GB on Amazon → Affiliate link — we may earn a commission on purchases, at no extra cost to you. Product picked by AI based on this article; it is not a tested recommendation.Source: tomshardware.com — Stephen Warwick, 2026-09-07
Published there as: “Astonishing mod runs DLSS 5 on a second GPU to boost neural-rendered FPS up to 127% — game renders on one card, neural post-processing runs on the other, much like dedicated PhysX GPUs”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.