9 papers
MPK: A Compiler and Runtime for Mega-Kernelizing Tensor Programs
Xinhao Cheng, Zhihao Zhang, Yu Zhou +17
We introduce Mirage Persistent Kernel (MPK), the first compiler and runtime system that automatically transforms multi-GPU model inference into a single high-performance mega-kerne…
Coral: Cost-Efficient Multi-LLM Serving over Heterogeneous Cloud GPUs
Yixuan Mei, Zikun Li, Zixuan Chen +5
The usage of large language models (LLMs) has grown increasingly fragmented, with no single model dominating. Meanwhile, cloud providers offer a wide range of mid-tier and older-ge…
Event Tensor: A Unified Abstraction for Compiling Dynamic Megakernel
Hongyi Jin, Bohan Hou, Guanjie Wang +18
Modern GPU workloads, especially large language model (LLM) inference, suffer from kernel launch overheads and coarse synchronization that limit inter-kernel parallelism. Recent me…
FlexLLM: Token-Level Co-Serving of LLM Inference and Finetuning with SLO Guarantees
Gabriele Oliaro, Xupeng Miao, Xinhao Cheng +9
Finetuning large language models (LLMs) is essential for task adaptation, yet today's serving stacks isolate inference and finetuning on separate GPU clusters -- wasting resources…
Towards Efficient Generative Large Language Model Serving: A Survey from Algorithms to Systems
Xupeng Miao, Gabriele Oliaro, Zhihao Zhang +4
In the rapidly evolving landscape of artificial intelligence (AI), generative large language models (LLMs) stand at the forefront, revolutionizing how we interact with our data. Ho…
Mirage: A Multi-Level Superoptimizer for Tensor Programs
Mengdi Wu, Xinhao Cheng, Shengyu Liu +7
We introduce Mirage, the first multi-level superoptimizer for tensor programs. A key idea in Mirage is Graphs, a uniform representation of tensor programs at the kernel, thread…