Unverified50% confidenceFactExact time
On two matching idle NVIDIA GPUs, models that fit on a single card, such as SDXL, Qwen-Image, and Wan2.2 5B, split half of each step's computation to the second card, and the officially stated speedup is about 1.5 to 1.8 times.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/6/2026
First Seen
Valid until: 2/4/2027
Sources
Related Entities
Related Claims
UnverifiedThe ultra-low VRAM version for 8GB graphics cards will make compromises in generation quality and speed, and the actual experience still falls short compared to a 24GB graphics card.70% similarUnverifiedGPU 硬件编码器(如 NVIDIA NVENC、Intel QSV、AMD AMF)速度可达软件编码的 5-10 倍,但同等码率下画质通常略逊于调优的软件编码69% similarUnverifiedOn NVIDIA GPUs, the geometric mean speedup of the full low-level backward pass is 1.15–1.81×.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1143494API
curl https://kongchang.com/api/v1/knowledge/claims/1143494MCP
get_claim(id=1143494)