Does not work on 8 GPUs

```
I have some problems running this code on 8 gpus. It crashed at the line: 
assert(same.size() == 3); in reducepipeline.cu

What steps will reproduce the problem?
1. get 8 k40 gpu, install them in 2 PCI buses. 4 for each.
2. train with 512 mini batch, data parallelism.


```

Original issue reported on code.google.com by `wangmeng...@gmail.com` on 3 Oct 2014 at 5:04
