Google Cloud expands the reach of Nvidia GPUs on Compute Engine to India

Google Cloud has expanded the reach of Nvidia GPUs on Compute Engine across eight regions worldwide including Brazil, India, Japan and Singapore. This expansion comes right after Google Cloud launched the public beta of NVIDIA Tesla T4 GPU. Nvidia Tesla T4 happens to be the company's latest GPU offering on GCP for AI architecture, after Nvidia K80, P4, P100, and V100.

Commenting on the Nvidia Tesla T4 public beta launch, here's what Google Cloud product manager Chris Kleban has to say:

“The T4 joins our NVIDIA K80, P4, P100, and V100 GPU offerings, providing customers with a wide selection of hardware-accelerated compute options. The T4 is the best GPU in our product portfolio for running inference workloads. Its high-performance characteristics for FP16, INT8, and INT4 allow you to run high-scale inference with flexible accuracy/performance tradeoffs that are not available on any other accelerator."

Also Read | The next big thing: How Amazon is harnessing India’s tech ability to build Alexa skills for the masses

Nvidia T4 GPUs are expected to manage workload balance over GPU, including high-performance computing tasks, deep learning training and inference, machine learning, data analytics and graphics. Nvidia T4 is fabricated using Nvidia's new Turing architecture. It offers Turing Tensor Cores and new RT Cores. It offers 16GB graphics. The Nvidia T4 can be accessed for as low as $0.29 (approximately Rs 20.60) per hour per GPU on Google Cloud Platform on Preemptible VMs.

“On-demand instances start at $0.95 per hour per GPU, with up to a 30 percent discount with sustained use discounts,” Kleban said.

Nvidia explains T4's Turing architecture offers real-time ray tracing to let a single GPU render realistic-looking 3D graphics. It has RT Cores that process ray tracing and are used by Nvidia's RTX ray-tracing technology, as well as APIs such as Microsoft DXR, Nvidia OptiX and Vulkan ray tracing.