David Gregg
Academic
Professor David Gregg is an associate professor of computer science and Fellow of Trinity College Dublin. His research deals with software performance optimization, particularly for multicore and low-power embedded systems. He has successfully commercialized outputs from his research, and he works closely with companies such as Movidius and IBM Research. He currently serves as Head of Software Systems within Trinity College.
.
2020. Beyond Base-2 Logarithmic Number Systems (WiP Paper). The 21st ACM SIGPLAN/SIGBED Conference on Languages, Compilers, and Tools for Embedded Systems.
.
2020. High-Performance Low-Memory Lowering: GEMM-based Algorithms for DNN Convolution. 2020 IEEE 32nd International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD).
.
2019. A Taxonomy of Channel Pruning Signals in CNNs. CoRR. abs/1906.04675
.
2019. No Room for Squares: Using Bitmap Masks to Improve Pedestrian Detection Using CNNs. IMVIP 2019:.
.
2018. Optimal DNN Primitive Selection with Partitioned Boolean Quadratic Programming. Proceedings of the 2018 International Symposium on Code Generation and Optimization.
.
2018. Scalar Arithmetic Multiple Data: Customizable Precision for Deep Neural Networks. CoRR. abs/1809.10572
.
2018. Low Complexity Multiply-Accumulate Units for Convolutional Neural Networks with Weight-Sharing. ACM Trans. Archit. Code Optim.. 15:31:1–31:24.
.
2018. Improving accuracy of Winograd convolution for DNNs. CoRR.
.
2017. Parallel Multi Channel convolution using General Matrix Multiplication. 2017 IEEE 28th International Conference on Application-specific Systems, Architectures and Processors (ASAP)2017 IEEE 28th International Conference on Application-specific Systems, Architectures and Processors (ASAP).
