Track manipulation, locomotion and embodied systems.
Search papers, labs, and topics across Lattice.
To resolve the severe latency and memory bottlenecks hindering 33-billion-parameter video diffusion models like MiniMax-H3, Sol-H3 introduces a full-stack acceleration framework spanning algorithmic scheduling to low-level operator search. The architecture pairs a cross-resolution denoising schedule—linked via a learned latent-to-latent mapping module that bypasses costly intermediate VAE roundtrips—with a hardware-aware Recursive Self-Improvement loop optimizing kernel fusions and memory layouts. These optimizations deliver up to a 30x speedup and 20% memory reduction, achieving 3.5x faster-than-real-time 1344x768 video generation on an 8xGB200 cluster and sub-minute execution on edge hardware.
Massive 33B video diffusion models can run 3.5x faster than real time by replacing VAE decode-reencode cycles with direct latent-to-latent resolution transfers and automated kernel synthesis.
Video diffusion models are rapidly scaling and exhibiting enhanced generation capabilities. Among these recent advancements, MiniMax-H3 stands out as a highly capable, production-level open-source model. However, its 33-billion parameters and multi-step iterative denoising process introduce substantial computational overhead. Consequently, their practical production is hindered by generation latency in the cloud deployment like NVIDIA-GB200, alongside strict memory limits that pose further challenges at the edge device like DGX-Spark. To address these diverse hardware bottlenecks from cloud to edge device, we present a full-stack inference pipeline that integrates efficient algorithmic design with optimized operator implementations. Algorithmically, we introduce a cross-resolution two-stage generation scheduler that exploits the step-wise nature of diffusion: early low-resolution steps rapidly establish the global layout, while later high-resolution steps focus refinements of local and perceptual details. These stages are connected by a learned latent-to-latent mapping module, completely eliminating the computationally expensive VAE decode-reencode cycle for resolution transferring cross different resolutions. For operator implementation, we deploy a Recursive Self-Improvement (RSI) loop that searches kernel fusions and memory layouts, evaluating latency together with numerical agreement. Together, these optimizations deliver up to 30x end-to-end speedup and 20% lower memory: a 5-second 1344x768 video with audio is generated 3.5x faster than real time on an 8xGB200 node, and in under a minute fully memory-resident on a single DGX Spark.
NVIDIA