4 ms·
If they get TensorFlow/Torch/MXNet running with no changes in the code, 32GB HBM2 for training models would be awesome! These days hitting memory limitations is
by bitL 8y ago
If they get TensorFlow/Torch/MXNet running with no changes in the code, 32GB HBM2 for training models would be awesome! These days hitting memory limitations is biting more than general slowness of GPUs during Deep Learning training.
- bcatanzaro 8y ago32GB HBM2 Nvidia Volta GPUs are available now, just in case you hadn’t heard.
- c2h5oh 8y agoHm? not close enough? https://github.com/ROCmSoftwarePlatform/tensorflow https://github.com/ROCmSoftwarePlatform/tensorflow https://github.com/ROCmSoftwarePlatform/cutorch_hip https://github.com/ROCmSoftwarePlatform/cutorch_hip https://github.com/ROCmSoftwarePlatform/mxnet https://github.com/ROCmSoftwarePlatform/mxnet https://github.com/ROCmSoftwarePlatform/hipCaffe https://github.com/ROCmSoftwarePlatform/hipCaffe
- infocollector 8y agoIs TF running on AMD Vega's? Keras?
- c2h5oh 8y ago> Is TF running on AMD Vega's? Not with mainline kernel (which supports up to Polaris), but it it already got merged into 4.18 that is going to be released in a couple of weeks. Until then it requires a bit extra effort https://github.com/RadeonOpenCompute/ROCm#dkms-driver-installation https://github.com/RadeonOpenCompute/ROCm#dkms-driver-instal... > Keras? over opencl with https://github.com/plaidml/plaidml https://github.com/plaidml/plaidml
- nl 8y agoThe problem is that it’s TF 1.3. TF is up to v1.8, and if you want to use any recent model you really need 1.4+. And even ignoring that, you still can’t use PyTorch.
- symisc_devel 8y agoWell, you can train your own models on the CPU for real-time object detection tasks using the SOD ML library we just released recently https://sod.pixlab.io/c_api/sod_realnet_train_start.html https://sod.pixlab.io/c_api/sod_realnet_train_start.html https://sod.pixlab.io https://sod.pixlab.io
- p1esk 8y agoHow much time does it take, compared to 1080Ti?
- symisc_devel 8y agoIt depends on your dataset. For example, a single class (face, pedestrian or car) object detection model with ten of thousands of positive samples should take at least 18 hours to complete on a modern Intel CPU with a target FPR (false positive rate) set to 0.00001. The lower the fpr is, the more training time is required. https://sod.pixlab.io/samples.html https://sod.pixlab.io/samples.html for some code samples.
- p1esk 8y agoI mean standard dataset and standard model, for example YOLO v2 on COCO. Say 1950X vs 1080Ti.
- WhitneyLand 8y agoSorry, I’m sure you didn’t mean to ignore the question. The commenter asked how the performance compares to an NVidia 1080Ti GPU. A precise benchmark may not be necessary but in general, it will take for example how many Skylake cores at 3.5 Ghz to approximate the GPU example?
- andbberger 8y agoOrders of magnitudes slower....