Research & Development Labs
Faculty in charge :
Dr. B Sathish Babu and Dr. Somesh Nandi
ML Frameworks:
TensorFlow, Keras, PyTorch, PyTorch Lightning, HuggingFace, PaddlePaddle, Halo and ONNX.
Deployment Options
Bare metal (Linux), VM (HyperV), Containers (Docker)
Host-Links
RDMA-based disaggregation between a host and IPU over a 100 Gbps RoCEv2 NIC, using the IPU over Fabric (IPUoF) protocol.
Host-to-IPU ratios supported: 1:16 up to 1:64
Graphcore Communication Library (GCL)
IPU-optimised communication and a collective library integrated with the Poplar SDK stack.
Support all-reduce (sum, max), all-gather, reduce, broadcast.
Scale near-linear performance to 64k IPUs.
PopVision
Visualisation and analysis tools.