AILatent Spaceabout 3 hours ago
Baseten Raises $13B, Inference Engineering Masterclass
110 min read
Baseten raised a $13B Series F, becoming an AI infrastructure decacorn. Philip Kiely and Ali Taha detailed inference techniques including cache-aware routing, quantization, and speculative decoding. Quantizing more of GLM-5.2 preserved quality while boosting throughput by 20%. Inference optimizations can deliver gains of 20%, 100%, or 200%.
Level
Hype check
Tap to vote and see what everyone thinks.
Summary by ByteBrief
