Google TurboQuant: The Algorithm That Cuts AI Memory Costs by 50% or More
Google's new TurboQuant algorithm achieves 8x memory reduction in AI models with zero accuracy loss, potentially transforming AI deployment economics.
Tag: LLM optimization
IndexCache: Tsinghua Researchers Achieve 1.82x Speed Boost for Long-Context AI Inference
Researchers at Tsinghua University have developed IndexCache, a new technique that achieves 1.82x faster inference for long-context AI models by eliminating redundant sparse attention computations.