FlashDecoding++: Faster Large Language Model Inference on GPUs

• URL: http://arxiv.org/abs/2311.01282v4
• Date: Fri, 5 Jan 2024 12:41:13 GMT
• Title: FlashDecoding++: Faster Large Language Model Inference on GPUs
• Authors: Ke Hong, Guohao Dai, Jiaming Xu, Qiuli Mao, Xiuhong Li, Jun Liu, Kangdi Chen, Yuhan Dong, Yu Wang
• Abstract summary: We present FlashDecoding++, a fast inference engine supporting mainstream Large Language Model (LLM) inference. To tackle the above challenges, FlashDecoding++ introduces a unified max value technique for different partial softmax computations. FlashDecoding++ can achieve up to 4.86x and 2.18x speedup on both NVIDIA and AMD GPUs.
• Score: 16.289377349637995
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: As the Large Language Model (LLM) becomes increasingly important in various domains. However, the following challenges still remain unsolved in accelerating LLM inference: (1) Synchronized partial softmax update. The softmax operation requires a synchronized update operation among each partial softmax result, leading to ~20% overheads for the attention computation in LLMs. (2) Under-utilized computation of flat GEMM. The shape of matrices performing GEMM in LLM inference is flat, leading to under-utilized computation and >50% performance loss after padding zeros in previous designs. (3) Performance loss due to static dataflow. Kernel performance in LLM depends on varied input data features, hardware configurations, etc. A single and static dataflow may lead to a 50.25% performance loss for GEMMs of different shapes in LLM inference. We present FlashDecoding++, a fast LLM inference engine supporting mainstream LLMs and hardware back-ends. To tackle the above challenges, FlashDecoding++ creatively proposes: (1) Asynchronized softmax with unified max value. FlashDecoding++ introduces a unified max value technique for different partial softmax computations to avoid synchronization. (2) Flat GEMM optimization with double buffering. FlashDecoding++ points out that flat GEMMs with different shapes face varied bottlenecks. Then, techniques like double buffering are introduced. (3) Heuristic dataflow with hardware resource adaptation. FlashDecoding++ heuristically optimizes dataflow using different hardware resource considering input dynamics. Due to the versatility of optimizations in FlashDecoding++, FlashDecoding++ can achieve up to 4.86x and 2.18x speedup on both NVIDIA and AMD GPUs compared to Hugging Face implementations. FlashDecoding++ also achieves an average speedup of 1.37x compared to state-of-the-art LLM inference engines on mainstream LLMs.

Related papers

• Pyramidal Flow Matching for Efficient Video Generative Modeling [67.03504440964564]
  This work introduces a unified pyramidal flow matching algorithm. It sacrifices the original denoising trajectory as a series of pyramid stages, where only the final stage operates at the full resolution. The entire framework can be optimized in an end-to-end manner and with a single unified Diffusion Transformer (DiT)
  arXiv  Detail & Related papers  (2024-10-08T12:10:37Z)
• vTensor: Flexible Virtual Tensor Management for Efficient LLM Serving [53.972175896814505]
  Large Language Models (LLMs) are widely used across various domains, processing millions of daily requests. Large Language Models (LLMs) are widely used across various domains, processing millions of daily requests.
  arXiv  Detail & Related papers  (2024-07-22T14:37:58Z)
• Enabling High-Sparsity Foundational Llama Models with Efficient Pretraining and Deployment [56.44025052765861]
  Large language models (LLMs) have revolutionized Natural Language Processing (NLP), but their size creates computational bottlenecks. We introduce a novel approach to create accurate, sparse foundational versions of performant LLMs. We show a total speedup on CPUs for sparse-quantized LLaMA models of up to 8.6x.
  arXiv  Detail & Related papers  (2024-05-06T16:03:32Z)
• Transformer-Lite: High-efficiency Deployment of Large Language Models on Mobile Phone GPUs [15.276687781165608]
  Large Language Model (LLM) is widely employed for tasks such as intelligent assistants, text summarization, translation, and multi-modality on mobile phones. To facilitate high-efficiency LLM deployment on device GPU, we propose four optimization techniques. We implement these techniques in our mobile inference engine, Transformer-Lite, which is compatible with both Qualcomm and MTK processors.
  arXiv  Detail & Related papers  (2024-03-29T08:26:53Z)
• Break the Sequential Dependency of LLM Inference Using Lookahead Decoding [27.87483106859749]
  Lookahead decoding is an exact, parallel decoding algorithm for large language models (LLMs) Our implementation can speed up autoregressive decoding by up to 1.8x on MT-bench and 4x with strong scaling on multiple GPUs in code completion tasks.
  arXiv  Detail & Related papers  (2024-02-03T06:37:50Z)
• FlightLLM: Efficient Large Language Model Inference with a Complete Mapping Flow on FPGAs [23.381331567339526]
  Transformer-based Large Language Models (LLMs) have made a significant impact on various domains. This paper proposes FlightLLM, enabling efficient LLMs inference with a complete mapping flow on FPGAs. FlightLLM beats NVIDIA A100 GPU with 1.2$times$ higher throughput using the latest Versal VHK158 FPGA.
  arXiv  Detail & Related papers  (2024-01-08T13:00:53Z)
• Efficient LLM inference solution on Intel GPU [19.154403468201924]
  Transformer based Large Language Models (LLMs) have been widely used in many fields. We propose an efficient LLM inference solution with low latency and high throughput. Compared with the standard HuggingFace implementation, the proposed solution achieves up to 7x lower token latency and 27x higher throughput.
  arXiv  Detail & Related papers  (2023-12-19T05:40:43Z)
• FusionAI: Decentralized Training and Deploying LLMs with Massive Consumer-Level GPUs [57.12856172329322]
  We envision a decentralized system unlocking the potential vast untapped consumer-level GPU. This system faces critical challenges, including limited CPU and GPU memory, low network bandwidth, the variability of peer and device heterogeneity.
  arXiv  Detail & Related papers  (2023-09-03T13:27:56Z)
• LM-Infinite: Zero-Shot Extreme Length Generalization for Large Language Models [83.98062659664785]
  Large language models (LLMs) typically train on short text segments (e.g., 4K tokens) due to the quadratic complexity of their Transformer architectures. This work identifies three major factors contributing to this length generalization failure. We propose LM-Infinite, a simple and effective method for enhancing LLMs' capabilities of handling long contexts.
  arXiv  Detail & Related papers  (2023-08-30T16:47:51Z)
• INR-Arch: A Dataflow Architecture and Compiler for Arbitrary-Order Gradient Computations in Implicit Neural Representation Processing [66.00729477511219]
  Given a function represented as a computation graph, traditional architectures face challenges in efficiently computing its nth-order gradient. We introduce INR-Arch, a framework that transforms the computation graph of an nth-order gradient into a hardware-optimized dataflow architecture. We present results that demonstrate 1.8-4.8x and 1.5-3.6x speedup compared to CPU and GPU baselines respectively.
  arXiv  Detail & Related papers  (2023-08-11T04:24:39Z)
• FlashAttention-2: Faster Attention with Better Parallelism and Work Partitioning [11.508362885430133]
  We exploit the asymmetric GPU memory hierarchy to bring significant memory saving and runtime speedup. FlashAttention is still not nearly as fast as optimized matrix-multiply (GEMM) operations, reaching only 25-40% of the theoretical maximum FLOPs/s. We propose FlashAttention-2, with better work partitioning to address these issues.
  arXiv  Detail & Related papers  (2023-07-17T17:50:36Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.