# GPNPU: Enabling Efficient Hardware-Based Direct Convolution with Multi-Precision Support in GPU Tensor Cores

> Research article (2020 57th ACM/IEEE Design Automation Conference (DAC), 2020) · cited 12× · AI/ML

**Wikidata**: [openalex:W3091804697](https://www.wikidata.org/wiki/openalex:W3091804697)  
**Source**: https://4ort.xyz/entity/gpnpu-enabling-efficient-hardware-based-direct-convolution-with-multi-precision-support-in-gpu-tensor-cores