Close Menu
    DevStackTipsDevStackTips
    • Home
    • News & Updates
      1. Tech & Work
      2. View All

      Sunshine And March Vibes (2025 Wallpapers Edition)

      May 31, 2025

      The Case For Minimal WordPress Setups: A Contrarian View On Theme Frameworks

      May 31, 2025

      How To Fix Largest Contentful Paint Issues With Subpart Analysis

      May 31, 2025

      How To Prevent WordPress SQL Injection Attacks

      May 31, 2025

      Windows 11 version 25H2: Everything you need to know about Microsoft’s next OS release

      May 31, 2025

      Elden Ring Nightreign already has a duos Seamless Co-op mod from the creator of the beloved original, and it’ll be “expanded on in the future”

      May 31, 2025

      I love Elden Ring Nightreign’s weirdest boss — he bargains with you, heals you, and throws tantrums if you ruin his meditation

      May 31, 2025

      How to install SteamOS on ROG Ally and Legion Go Windows gaming handhelds

      May 31, 2025
    • Development
      1. Algorithms & Data Structures
      2. Artificial Intelligence
      3. Back-End Development
      4. Databases
      5. Front-End Development
      6. Libraries & Frameworks
      7. Machine Learning
      8. Security
      9. Software Engineering
      10. Tools & IDEs
      11. Web Design
      12. Web Development
      13. Web Security
      14. Programming Languages
        • PHP
        • JavaScript
      Featured

      Oracle Fusion new Product Management Landing Page and AI (25B)

      May 31, 2025
      Recent

      Oracle Fusion new Product Management Landing Page and AI (25B)

      May 31, 2025

      Filament Is Now Running Natively on Mobile

      May 31, 2025

      How Remix is shaking things up

      May 30, 2025
    • Operating Systems
      1. Windows
      2. Linux
      3. macOS
      Featured

      Windows 11 version 25H2: Everything you need to know about Microsoft’s next OS release

      May 31, 2025
      Recent

      Windows 11 version 25H2: Everything you need to know about Microsoft’s next OS release

      May 31, 2025

      Elden Ring Nightreign already has a duos Seamless Co-op mod from the creator of the beloved original, and it’ll be “expanded on in the future”

      May 31, 2025

      I love Elden Ring Nightreign’s weirdest boss — he bargains with you, heals you, and throws tantrums if you ruin his meditation

      May 31, 2025
    • Learning Resources
      • Books
      • Cheatsheets
      • Tutorials & Guides
    Home»Development»Machine Learning»This AI Paper from ByteDance Introduces MegaScale-Infer: A Disaggregated Expert Parallelism System for Efficient and Scalable MoE-Based LLM Serving

    This AI Paper from ByteDance Introduces MegaScale-Infer: A Disaggregated Expert Parallelism System for Efficient and Scalable MoE-Based LLM Serving

    April 9, 2025
    This AI Paper from ByteDance Introduces MegaScale-Infer: A Disaggregated Expert Parallelism System for Efficient and Scalable MoE-Based LLM Serving

    Large language models are built on transformer architectures and power applications like chat, code generation, and search, but their growing scale with billions of parameters makes efficient computation increasingly challenging. Scaling such systems while maintaining low latency and high throughput puts pressure on algorithm design and system-level optimization. Effectively serving these models now requires careful orchestration of memory, communication, and compute resources.

    A critical challenge in this space is how sparsity, introduced through Mixture-of-Experts (MoE) models, affects inference performance. These models selectively activate a subset of feed-forward networks per input, reducing computational load. However, this selective activation leads to underutilization of hardware. During inference, attention modules become bottlenecks due to frequent memory access to key-value caches, while the FFN modules become idle because each receives a small fraction of tokens. As a result, GPU utilization drops significantly, especially during decoding, creating inefficiencies and inflating operational costs.

    While some methods like vLLM and TensorRT-LLM have attempted to address inference scaling through parallelism and optimized kernels, these solutions remain constrained. They process the model holistically, meaning they cannot independently adjust scaling for different components. As MoE models grow in size and sparsity, this approach leads to smaller active batches per expert, weakening the benefits of batching for FFNs. Moreover, tensor and pipeline parallelism approaches add communication overhead, especially across nodes, which becomes a limiting factor in multi-GPU environments.

    ByteDance and Peking University researchers have introduced MegaScale-Infer, a system that rethinks the architecture of MoE serving. Instead of serving the model as a monolithic block, the researchers disaggregate the attention and FFN modules, deploying them on separate GPUs. This separation enables customized scaling and parallelism strategies tailored to the specific needs of each module. Attention modules, which are memory-intensive, are replicated to aggregate requests, while FFN modules are scaled using expert parallelism. The system also supports heterogeneous GPU deployment, assigning cost-effective memory-heavy GPUs to attention tasks and compute-optimized GPUs to FFNs. This disaggregation dramatically improves resource usage and flexibility in deployment.

    To further optimize performance, MegaScale-Infer employs a ping-pong pipeline parallelism strategy. The idea is to break down batches of requests into smaller micro-batches that alternate between attention and FFN modules, ensuring that neither component sits idle. The system determines the optimal number of micro-batches required to maintain high utilization, considering compute time, communication latency, and hardware setup. For example, if the communication time is less than half the compute time, at least three micro-batches are used. Further, the system integrates a high-performance M2N communication library that avoids unnecessary GPU-to-CPU data copies, reducing latency and instability. This library replaces the traditional All-to-All routing with a more efficient sender-receiver model designed specifically for MoE’s token dispatch pattern.

    MegaScale-Infer was tested on multiple large-scale MoE models, including Mixtral 8×22B, DBRX, and a scaled custom model with 317 billion parameters. In experiments on homogeneous setups using NVIDIA Ampere GPUs, MegaScale-Infer improved per-GPU decoding throughput by up to 2.56× compared to vLLM and 1.28× over TensorRT-LLM. The scaled model achieved a 7.11× gain over vLLM and a 1.90× gain over TensorRT-LLM. On heterogeneous clusters with H20 GPUs for attention and L40S for FFNs, the system achieved up to 3.24× and 1.86× higher throughput per dollar than the baselines. Its M2N communication library delivered up to 4.2× higher throughput and 68.2% lower latency than NCCL.

    This paper presents a clear problem of underutilized GPUs during MoE inference and offers a practical solution by modularizing the architecture. The proposed disaggregation strategy, combined with micro-batch pipelining and a custom communication protocol, substantially impacts serving efficiency and cost.


    Check out the Paper. All credit for this research goes to the researchers of this project. Also, feel free to follow us on Twitter and don’t forget to join our 85k+ ML SubReddit.

    🔥 [Register Now] miniCON Virtual Conference on OPEN SOURCE AI: FREE REGISTRATION + Certificate of Attendance + 3 Hour Short Event (April 12, 9 am- 12 pm PST) + Hands on Workshop [Sponsored]

    The post This AI Paper from ByteDance Introduces MegaScale-Infer: A Disaggregated Expert Parallelism System for Efficient and Scalable MoE-Based LLM Serving appeared first on MarkTechPost.

    Source: Read More 

    Facebook Twitter Reddit Email Copy Link
    Previous ArticleHuawei Noah’s Ark Lab Released Dream 7B: A Powerful Open Diffusion Reasoning Model with Advanced Planning and Flexible Inference Capabilities
    Next Article Understanding the :root Selector and CSS Variables

    Related Posts

    Machine Learning

    How to Evaluate Jailbreak Methods: A Case Study with the StrongREJECT Benchmark

    May 31, 2025
    Machine Learning

    Cisco’s Latest AI Agents Report Details the Transformative Impact of Agentic AI on Customer Experience

    May 31, 2025
    Leave A Reply Cancel Reply

    Continue Reading

    jeremy379/laravel-openid-connect

    Development

    Downgrade Attacks Could Affect Fully Updated Windows Systems With Previously Patched Vulnerabilities

    Development

    NVIDIA’s RTX 4060 is aging horribly as Indiana Jones and the Great Circle Nazi-punches your VRAM

    Development

    Huawei Research Developed MatMulScan: A Parallel Scan Algorithm Transforming Parallel Computing with Tensor Core Units, Enhancing Efficiency and Scalability for Large-Scale Matrix Operations

    Development

    Highlights

    Figma’s AI design feature disabled after copying Apple’s weather app

    July 5, 2024

    Figma CEO Dylan Field announced that his company would temporarily disable its new “Make Design”…

    I’ve wanted to play this upcoming Xbox exclusive since its 2021 reveal, and I finally got my chance

    June 22, 2024

    The best free antivirus software of 2024: Expert tested

    August 22, 2024

    Woman scammed out of €800k by an AI deep fake of Brad Pitt

    January 15, 2025
    © DevStackTips 2025. All rights reserved.
    • Contact
    • Privacy Policy

    Type above and press Enter to search. Press Esc to cancel.