GPU Build
Jump to navigation
Jump to search
| GPU Build | |
|---|---|
| Project Information | |
| Project Title | GPU Build |
| Owner | Oliver Chang, Kyran Adams |
| Start Date | |
| Deadline | |
| Primary Billing | |
| Notes | |
| Has project status | Active |
| Copyright © 2016 edegan.com. All Rights Reserved. | |
GPU
- "I quickly found that it is not only very difficult to parallelize neural networks on multiple GPUs efficiently, but also that the speedup was only mediocre for dense neural networks. Small neural networks could be parallelized rather efficiently using data parallelism, but larger neural networks... received almost no speedup."
- Possible other use of multiple GPUs: training multiple different models simultaneously, "very useful for researchers, who want try multiple versions of a new algorithm at the same time."
- This source recommends GTX 1080 Tis and does cost analysis of it
- If the network doesn't fit in the memory of one GPU (11 GB),
- Might want to get two graphics cards, one for development, one (crappy card) for operating system
- Intra-model parallelism: If a model has long, independent computation paths, then you can split the model across multiple GPUs and have each compute a part of it. This requires careful understanding of the model and the computational dependencies.
- Replicated training: Start up multiple copies of the model, train them, and then synchronize their learning (the gradients applied to their weights & biases).
TL;DR
Using multiple GPUs adds a lot of complexity. It has a few benefits: possible speed ups if the network can be split up (and is big enough), able to train multiple networks at once (either copies of the same network or modified networks), more memory for huge batches. Some frameworks have much better performance with multiple GPUs (pytorch, caffe 2) while others are catching up.
Todo: ask what software, dataset size, development approach
