Machine LearningChips & Compute40 min reading time

Presentation: From S3 to GPU in One Copy: Rethinking Data Loading for ML Training

InfoQ (AI, ML & Data)
Read full post
SpiralDB's Onur Satici demonstrates how their open-source Vortex file format enables high-speed data loading from S3 storage directly to GPUs, optimizing bandwidth by column pruning. This approach reduces GPU idle time and improves efficiency in machine learning training workflows.

More in Machine Learning

Machine Learning4 min read

DeepSeek launches V4.1-Flash and retires V4-Pro, its flagship model

Covered by 2 sources
Machine Learning4 min read

Mistral wants open-weight AI to compete at the frontier. It just raised $3.5 billion to do it.

The New Stack (AI)
Machine Learning4 min read

Salesforce introduces Enterprise AI Harness, AI Control Plane

SiliconANGLE