I have been running Caffe on the TK1 and I have noticed that I get hardly any improvement in performance with the forward or backward pass times when linking to cuDNN v2. The improvement is around a factor 1.05x better than without cuDNN. Is this to be expected for cuDNN on the TK1 ? Has anyone seen better acceleration ?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Performance of Caffe with cuDNN on the TK1 | 3 | 1293 | March 14, 2016 | |
| cuDNN 5 on Tegra K1 | 2 | 1673 | June 24, 2016 | |
| Caffe and Imagenet | 3 | 1878 | March 17, 2015 | |
| ANNOUNCING NVIDIA® cuDNN EMBEDDED – GPU Accelerated Machine Learning for Jetson TK1 | 1 | 2229 | December 17, 2014 | |
| ANNOUNCING NVIDIA® cuDNN EMBEDDED – GPU Accelerated Machine Learning for Jetson TK1 | 1 | 2746 | April 22, 2017 | |
| cuDNN v2: Higher Performance for Deep Learning on GPUs | 2 | 453 | November 18, 2015 | |
| Cudnn may be slower? | 3 | 2643 | September 28, 2015 | |
| Cudnn8 regression on TX2 | 6 | 1063 | October 18, 2021 | |
| Slow convolution speeds on TK1 | 2 | 587 | February 2, 2018 | |
| Tegra K1 vs Tegra X1 vs Tegra X2 cuDNN convolution speeds | 0 | 946 | January 15, 2018 |