i'd think the 'gpu intensive ones can literally be a different 'class' of apps. e.g. Tensorflow and or Caffe/Torch
and those apps generates the cpu based algorithms optimizations for rosetta@home itself. i.e. Tensorflow and or Caffe/Torch does all the deep learning data mining and generate likely folds and r@h explore the folds and confirm the folds for unknown molecules / proteins

Today you can "combine" different approach to bionformatical problems, like this:

The application software programs used for the pipeline include:

AutoDock Vina, a molecular docking program
NAMD, a molecular dynamics program
TensorFlow, a machine learning toolkit

Each compute node includes one AMD EPYC processor, 64 GB of memory, one InfiniBand adaptor, and four Radeon Instinct™ MI25 Accelerators, which combine AMD GPU technologies and the ROCm software platform

Cpu+gpu+Ai
We've got some interesting times ahead

---

Is there anyone who could point out address for CPU version of source code? I may be able to port the code into OpenCL / Cuda.

Is there anyone who could point out address for CPU version of source code?

If you look up the posts by "rjs5", you might get some info.
He has been working with the Rosetta developers for some time (on AVX, etc.), but I think most of it is proprietary.

---

Is there anyone who could point out address for CPU version of source code?

The code is accessible for researchers, but you have to request access to admins

---

Is there anyone who could point out address for CPU version of source code? I may be able to port the code into OpenCL / Cuda.

If you want, there is Tn-Grid.
They have not a gpu developer and are interested in it
And they have open source code

---

If you want, there is Tn-Grid.
They have not a gpu developer and are interested in it

Also, if you look on page 23 of their very interesting PowerPoint presentation, they mention the possibility of a GPU.
http://gene.disi.unitn.it/test/documents/Gene@Home.@BF15.pdf

---

Also, if you look on page 23 of their very interesting PowerPoint presentation, they mention the possibility of a GPU.
http://gene.disi.unitn.it/test/documents/Gene@Home.@BF15.pdf

I have often talked with ValterC, the project admin, and he said that they had not this kind of developer.
But they are open to collaboration with volunteers (in the past the avx/sse app is developed with the contribution of a volunteer).

---

You will be happy to know that I have ordered a second RX 570. So I am good for either CUDA or OpenCl.

---

You will be happy to know that I have ordered a second RX 570. So I am good for either CUDA or OpenCl.

Amd RX570? Only OpenCl...

---

I have a bunch of Nvidia cards, used mainly for Folding.

---