Home▸Physics & Mechanics▸Pipeline Parallelism Bubble Lab: GPU Stages, Micro-batches and Idle Time

🧮 Pipeline Parallelism Bubble Lab: GPU Stages, Micro-batches and Idle Time

Pipeline-parallel training of one model across p GPU stages with m micro-batches: GPipe and 1F1B schedules simulated as a dependency graph, bubble fraction (p-1)/(m+p-1), makespan (m+p-1)(tf+tb), peak stored activations m versus min(p,m), with Blender-built GPU cards.

Physics & Mechanics3DModerate60 FPS📱 Mobile-adapted🟠 Blender⇄ 2D version
Pipeline Parallelism Bubble Lab: GPU Stages, Micro-batches and Idle Time ↗ Open standalone
⚙ Under the hood

Pipeline-parallel training of one model across p GPU stages with m micro-batches: GPipe and 1F1B schedules simulated as a dependency graph, bubble fraction (p-1)/(m+p-1), makespan (m+p-1)(tf+tb), peak stored activations m versus min(p,m), with Blender-built GPU cards.

pipeline parallelismGPipe1F1Bpipeline bubbledistributed trainingmicro-batchGPUBlender

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install

What did you find?

Add reproduction steps (optional)