Oracle and Nvidia partner on analytics and AI
October 18, 2018
At last week’s GPU Technology Conference in Munich, Oracle announced that it was the first public cloud provider to support the Nvidia HGX-2 platform to meet the needs of analytics, machine learning and AI.
The companies also announced the general availability of support for GPU-accelerated deep learning and HPC containers from the Nvidia GPU Cloud (NGC) container registry on Oracle’s cloud infrastructure.
"We are very excited about this collaboration with Nvidia," said Clay Magouyrk, senior vice president at Oracle. “As the world of computing continues to push the boundaries of what's possible, we are providing our customers with the software, tools and cloud infrastructure needed to solve the most complex challenges. Whether you are an engineer, data scientist, researcher or developer, we are bringing the power of compute and cloud to your fingertips.”
From enabling autonomous vehicles to driving global climate simulations, rapid progress in AI and HPC has transformed entire industries while also demanding massive increases in complexity and compute power. HGX-2 is designed for multi-precision computing to accelerate the most demanding applications by unleashing two petaflops of computing power and half a terabyte of total GPU memory with 16 Nvidia Tesla V100 Tensor core GPUs interconnected with NVSwitch.
Supporting HGX-2 on both Oracle’s cloud infrastructure bare-metal and virtual machine instances, Oracle and Nvidia are helping users solve AI and HPC problems for complex workloads.
California-based Oracle also announced support for Rapids open-source software introduced by Nvidia at the show for executing end-to-end data science training pipelines accelerated on its GPUs. Rapids is available on Oracle cloud infrastructure via NGC. It can accelerate data science pipelines by moving workflows onto the GPU. This optimises machine learning training with more iterations for better model accuracy.
Data scientists can integrate Rapids with few code changes, letting them accelerate the Python data science toolchain. With this offering and support for NGC containers, Oracle and Nvidia are helping users deploy containerised applications and frameworks for HPC, data science and AI and run them seamlessly on Oracle cloud infrastructure.
Oracle is also working with California-based Nvidia to support Rapids across its platform, including the Oracle data science cloud, to accelerate end to-end data science workflows. The software runs seamlessly on the Oracle cloud, so users can support their HPC, AI and data science needs, all while taking advantage of the portfolio of GPU instances available on Oracle’s cloud infrastructure.
"This new collaboration with Oracle will help fuel incredible innovation across a wide range of industries," said Ian Buck, vice president at Nvidia. "By taking advantage of Nvidia's latest technologies, Oracle is well positioned to meet surges in demand for GPU acceleration for deep learning, high-performance computing, data analytics and machine learning."