Eliminating GPU Starvation with tf.data Prefetching and Interleaving
Learn how to stop GPU starvation in TensorFlow by implementing tf.data prefetching and parallel interleaving to overlap data preprocessing with model execution.
Loading…
ReadMeFeed / Community knowledge
#gpu-optimization
Real questions. Useful conversations. Find the people who know your stack.
tensorflowOptimizing GPU Memory with Three.js BufferGeometry Attributes
Learn how to optimize Three.js performance using BufferGeometry attributes, typed arrays, and InterleavedBuffers to reduce GPU memory overhead and improve cache locality.
threejsOverlap CPU Data Loading with GPU Training Using tf.data Prefetch
End every tf.data training pipeline with .prefetch(tf.data.AUTOTUNE) to overlap CPU data loading with GPU training. This guide shows the correct transformation order, a worked TFRecord example, and common mistakes that silently break overlap.
Push Constants in Vulkan: Fast Per‑Draw Uniform Updates Without Descriptor Sets
Push constants let you update up to 128 bytes of shader data directly inside a Vulkan command buffer, eliminating descriptor sets for small, per‑draw uniforms. This article explains how to declare, use, and verify push constants, with a concrete example and practical checklist.
Optimizing Real-Time Visuals in Processing with the P2D Renderer
Learn how to optimize real-time visual synthesis in Processing using the P2D renderer, focusing on reducing PCIe bottlenecks and OpenGL state changes.
processingStop GPU Starvation: Optimizing Input Pipelines with tf.data
Stop letting your CPU hold back your GPU. Learn how to use tf.data's prefetch, cache, and parallel mapping to eliminate GPU starvation and speed up training.
Stop Using glTexImage2D: Moving to Immutable Texture Storage
Stop using glTexImage2D for every texture. Learn how glTexStorage2D reduces driver overhead, prevents texture incompleteness, and optimizes GPU memory layout.
Reducing Draw Call Overhead in libGDX with SpriteBatch and TextureAtlas
Learn how to optimize libGDX 2D rendering by leveraging SpriteBatch and TextureAtlas to minimize GPU draw calls and prevent performance‑killing texture switches.
libgdxManaging GPU Memory in PixiJS: Texture Cache Architecture
Learn how to optimize GPU memory in PixiJS by decoupling BaseTextures from Textures to prevent redundant VRAM uploads and memory leaks.
pixijs