Our great sponsors
-
InfluxDB
Power Real-Time Data Analytics at Scale. Get real-time insights from all types of time series data with InfluxDB. Ingest, query, and analyze billions of data points in real-time with unbounded cardinality.
Yes I definitely want to add GPU support, and totally agree that GPU support is very much needed. You can see the discussion so far here: https://github.com/coreylowman/dfdx/issues/9. Currently need to get access to a GPU machine to develop on for this...
In my opinion one of the most promising paths to replicating the Python/C++/CUDA deep learning ecosystem is the Triton compiler, which makes writing efficient kernel much much simpler than with CUDA and which can be embedded in other languages. It currently only supports Python, but there was some activity at one point around integrating Triton with Rust at one point.
Related posts
- Dfdx: Shape Checked Deep Learning in Rust
- rapl: Rank Polymorphic array library for Rust.
- Announcing cudarc and fully GPU accelerated dfdx: ergonomic deep learning ENTIRELY in rust, now with CUDA support and tensors with mixed compile and runtime dimensions!
- Tensor shapes with both const generic and run time dimensions
- Oxidizing Machine Learning