-
Multi-GPU operation and data / model / hybrid parallelism are planned and in development for Caffe. The purpose of the thread is to focus the conversation, since this has been asked here, there, and e…
-
I think Intel Caffe is using MLSL for distributed processing.
Is there any paper or report about how Intel Caffe conducts multi-node processing?
E.g. the way of communication, data partition, co…
-
Are Web Assembly specified in terms of [WebWorkers](https://developer.mozilla.org/en-US/docs/Web/API/Web_Workers_API/Using_web_workers), or are they different? What are the differences, or are WebWork…
-
Hi team,
Currently I'm saturating my GPU, which is cool. But on the same server I can have more GPU's - eg. an Amazon g2.8xlarge instance has 4 Nvidia K520's. I want to be able to use them too.
I wa…
-
I'm using the dynamic scheduler and I want to know the best way to order my work. I have about 400 units of work, each unit of work is a NxN matrix inversion (200-1000) elements. If I order staticly…
-
dask creates more processes than necessary during initialization in
https://github.com/dask/dask/blob/master/dask/threaded.py#L22 .
```
default_pool = ThreadPool()
```
On a system with 64 os …
-
Using the configure scripts for Sandy Bridge processors, we compiled both ELPA and Quantum ESPRESSO v.6.0, and see that pw.x builds perfectly well. However, when building the neb.x code, the executab…
-
I've identified what may be a serious bug in HOMME. (It is obviously causing errors in diagnostic output, but the issue could have much wider consequences.) There is a simple workaround however, so we…
-
The seed in computed recursively throughout the code which makes it inconvenient for parallelism.
I tried finding a function that could give me the value of the seed at any iteration.
F(x+1)=a*F(x)+…
-
**Originally by _dougmill AT us DOT ibm DOT COOM _ on 2010-01-20 09:27:01 -0600**
---
Attached is a document to provide background for the tickets discussing horizontal parallelism extensions to th…