E2E Networks launches managed NVIDIA Tesla V100 GPU compute instances for cloud customers.

PreferredSource E2E Networks, a leading Indian Cloud Provider, announced the launch of datacenter-grade Cloud GPU instances based on Nvidia's Tesla V100 via its Public Cloud. These instances are designed to support companies and startups where data scientists and engineers work on AI and ML workloads.

The Tesla V100 GPUs feature 32 GB of onboard graphics memory and are ideal for a wide range of machine learning and deep learning applications. These include Natural Language Processing and structured data analytics, Convolutional Neural Networks for image recognition and generation, Deep Analytics, Computer Vision, and Conversational Speech Recognition, among other use cases.

The GPU instances deliver bare-metal performance through Passthrough mode by being directly attached to Virtual Compute Nodes. They support multiple AI, ML, and deep learning frameworks including CUDA, TensorFlow, MXNet, Caffe2, OpenFOAM, Theano, and PyTorch.

E2E Networks' GPU instances can optimize operational costs by as much as 70 per cent compared to other leading Public Cloud providers. The instances are available as hourly billed options or pre-committed instances at deeply discounted pricing. All GPU instances are available from Indian data centres, ensuring data locality for critical India-centered data.