Lancium Compute Resources
Available Applications
Applications packaged to run on Lancium Compute contain all of the necessary prerequisite libraries. You only need to upload your data files along with the script you'd like to run to process them.
Our currently supported applications and their installed versions are:
TensorFlow
on Python 2
- Python 2.7
- Tensorflow 1.14
- Numpy 1.16
- Scipy 1.2
on Python 3
- Python 3.7
- Tensorflow 1.13
- Numpy 1.16
- Scipy 1.2
PyTorch
on Python 2
- Python 2.7
- PyTorch 1.1
- Numpy 1.16
on Python 3
- Python 3.7
- PyTorch 1.0
- Numpy 1.16
Theano
on Python 2
- Python 2.7
- Theano 1.0
- Numpy 1.16
- Scipy 1.2
on Python 3
- Python 3.7
- Theano 1.0
- Numpy 1.16
- Scipy 1.3
Caffe
on Python 2
- Python 2.7
- Caffe 1.0
- Numpy 1.16
- Scipy 1.2
- Pandas 0.24
- Scikit-image 0.14
on Python 3
- Python 3.7
- Caffe 1.0
- Numpy 1.16
- Scipy 1.2
- Pandas 0.24
- Scikit-image 0.15
Keras
on Python 2
Not currently packaged. Contact us if you'd like to see it available.
on Python 3
- Python 3.7
- Keras 2.2
- Tensorflow 1.13
- Numpy 1.16
- Scipy 1.2
If you have other frameworks you'd like packaged as a Lancium Compute application, or would like to see additional libraries in an existing application, let us know.
Compute Nodes
Lancium Compute nodes generally have the following specs. For details on the exact specifications of each performance tier, please check our pricing.
- Celeron G3930 or Xeon E5-2680 CPUs
- 2-16 CPU cores
- 16, 32, or 64 GB of memory
- NVIDIA 1070, K40, and K80 GPUs