VUDA: Enabling Controlled Spatial Sharing of Graphics and Compute on NVIDIA GPUs
cs.OS, cs.AI, cs.DC
Submitted: 2026-05-02
Updated: 2026-09-26
Code: https://github.com/NVIDIA/open-gpu-kernel-modules
Terminology
Sources
- WorldVLA: Towards Autoregressive Action World Model
- RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation
- WoVR: World Models as Reliable Simulators for Post-Training VLA Policies with RL
- Pagoda: An Energy and Time Roofline Study for DNN Workloads on Edge Accelerators
- Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
- Isaac Lab: A GPU-Accelerated Simulation Framework for Multi-Modal Robot Learning
- Proximal Policy Optimization Algorithms
- Serving DNN Models with Multi-Instance GPUs: A Case of the Reconfigurable Machine Scheduling Problem
- RLinf-VLA: A Unified and Efficient Framework for Reinforcement Learning of Vision-Language-Action Models
- Absence of percolation for infinite Poissonian systems of stopped paths
Related papers
- MemSpec: Memory-Aware Runtime for Adaptive Draft Scheduling in Speculative Decoding on Edge Devices
- Dynamic Flow, Static Graph: KV Cache Reuse for Efficient LLM Serving on Mobile NPUs
- Vulcan: Instance-specialized, Verifiable Systems Heuristics Through LLM-driven Search
- Planarian: Managing Agent State with Statepoints
- ActKV: Efficient LLM Agents through Action-Guided KV Cache Management
- GroupKV: Hierarchical KV Cache Management for Long-Context Diffusion LLM Inference