# 7.7 LNPU: A 25.3TFLOPS/W Sparse Deep-Neural-Network Learning Processor with Fine-Grained Mixed Precision of FP8-FP16

> Research article (2019 IEEE International Solid-State Circuits Conference - (ISSCC), 2019) · cited 155× · AI/ML

**Wikidata**: [openalex:W2921918777](https://www.wikidata.org/wiki/openalex:W2921918777)  
**Source**: https://4ort.xyz/entity/7-7-lnpu-a-25-3tflops-w-sparse-deep-neural-network-learning-processor-with-fine-grained-mixed-precision-of-fp8-fp16