15.1b East-West traffic: GPU-to-GPU gradient synchronization — bandwidth-sensitive

📦 AI Data Center Networking 📖 The AI Traffic Problem

🌐 Context Introduction

When training large AI models across multiple GPUs, the GPUs must constantly share information to stay synchronized. This communication happens between GPUs (not between a GPU and a storage server), which is called East-West traffic. The most critical type of East-West traffic is gradient synchronization — where GPUs exchange small but frequent updates during training. This traffic is extremely bandwidth-sensitive, meaning the speed of your network directly impacts how fast your model trains.


⚙️ What Is Gradient Synchronization?

  • During distributed training, each GPU computes its own set of gradients (mathematical adjustments to the model).
  • These gradients must be averaged across all GPUs before the model can be updated.
  • This averaging process is called all-reduce or gradient synchronization.
  • The traffic pattern is many-to-many: every GPU talks to every other GPU simultaneously.

📊 Why Is It Bandwidth-Sensitive?

Factor Explanation
Frequency Gradients are exchanged after every training step (often hundreds of times per second)
Size Each gradient exchange can be hundreds of megabytes to gigabytes
Burst nature All GPUs send data at the exact same moment, creating traffic spikes
Synchronization Training stalls until all GPUs finish exchanging — slow network means idle GPUs

🕵️ The East-West Traffic Challenge

  • North-South traffic (GPU to storage) is less time-critical — data can be buffered.
  • East-West traffic (GPU to GPU) must complete before the next training step can begin.
  • If the network is too slow, GPUs sit idle waiting for gradients — this is called network tail latency.
  • A single slow link can delay all GPUs in the cluster.

📊 Visual Representation: East-West Intra-Cluster Node-to-Node Traffic

This diagram displays East-West data paths connecting multiple compute nodes in a parallel processing environment.

flowchart LR Node1["Server Node 1"] -->|East-West| Node2["Server Node 2"] Node2 -->|East-West| Node3["Server Node 3"] Node3 -->|East-West| Node1 classDef cpu fill:#eafaf1,stroke:#76b900,stroke-width:2px,rx:6px,ry:6px; classDef memory fill:#f0f7ff,stroke:#3498db,stroke-width:1.5px,rx:4px,ry:4px; classDef system fill:#f1f5f9,stroke:#64748b,stroke-width:1.5px; class Node1,Node2,Node3 cpu;

🛠️ How Engineers Address Bandwidth Sensitivity

  • Use high-bandwidth interconnects: NVIDIA NVLink, InfiniBand (200 Gbps or 400 Gbps), or high-speed Ethernet (RoCEv2).
  • Minimize hops: Place GPUs that communicate frequently on the same switch or within the same rack.
  • Optimize topology: Use a fat-tree or Dragonfly topology to avoid oversubscription.
  • Enable RDMA (Remote Direct Memory Access): Allows GPUs to read/write each other's memory directly without involving the CPU — reduces latency.
  • Tune collective algorithms: Use ring all-reduce or tree all-reduce to balance traffic across links.

🔍 Real-World Example

Imagine training a large language model on 64 GPUs across 8 servers (8 GPUs per server).

  • After each training step, all 64 GPUs must exchange gradients.
  • Each GPU sends ~500 MB of gradient data to every other GPU.
  • Total traffic per step: 64 × 500 MB = 32 GB moving across the network.
  • With a 200 Gbps InfiniBand link, this takes ~1.3 seconds.
  • With a 25 Gbps Ethernet link, this takes ~10 seconds — GPUs are idle 8.7 seconds longer per step.

✅ Key Takeaways for New Engineers

  • East-West gradient sync is the bottleneck in most AI training clusters.
  • Bandwidth matters more than latency for this traffic pattern — you need fat pipes.
  • Network oversubscription (sharing bandwidth among many GPUs) kills training performance.
  • Always check your network bandwidth when troubleshooting slow training — it's often the hidden culprit.
  • Use tools like ib_write_bw or nccl-tests to measure actual GPU-to-GPU bandwidth in your cluster.

📚 Summary

Gradient synchronization is the heartbeat of distributed AI training. Every training step depends on all GPUs quickly exchanging their gradients. If the East-West network is too slow, your expensive GPUs spend more time waiting than computing. As an engineer, your job is to ensure the network is fast, non-blocking, and dedicated to this traffic — because in AI infrastructure, bandwidth is time, and time is money.