NVIDIA's new GPU-acceleration platform for machine learning

Doris Richards
October 14, 2018

This time though, Nvidia isn't announcing a new Graphics Processing Unit (GPU) platform, or a new proprietary SDK for deep learning, but is instead announcing new a set of new open source libraries for GPU-accelerated analytics and machine learning (ML). "Apache Arrow and in memory data format and some other tools that allow us to scale from using just one GPU to multiple GPUs in the system, to multiple node and clusters of GPUs", said Jeff Tseng, head of product for AI infrastructure at Nvidia, in a pre-announcement conference call. "It's a logical next step for us".

RAPIDS builds on popular open-source projects - including Apache Arrow, pandas and scikit-learn - by adding GPU acceleration to the most popular Python data science toolchain.

MapRs work with NVIDIA in the RAPIDS ecosystem is helping make broad adoption in the enterprise easy for the largest breadth of workloads, said Clment Farabet, vice president of AI infrastructure at NVIDIA.

"For 35 years, we have relied on Moore's Law". "Data analytics and machine learning is now the leading high-performance computing segment". "Moore's Law has ended at a time when demand for computing enhancement continues to grow".

According to the RAPIDS developers, by virtue of GPU acceleration, performance on these types of applications can be sped up by an order of magnitude or more - up to 50x on some workloads. Besides NVIDIA, the initial software was developed in concert with key open source providers, including Anaconda, BlazingDB, Databricks, Quansight and scikit-learn.

Additionally, some of the world's leading technology companies are supporting RAPIDS through new systems, data science platforms and software solutions, including IBM, Cisco, Dell EMC, Lenovo, NERSC, NetApp, Pure Storage, SAP and SAS.

Although in its current form, RAPIDS is built atop CUDA, the software suite is otherwise independent from NVIDIA and could presumably be targeted to AMD GPUs or other accelerators. "The world's largest industries run algorithms written by machine learning on a sea of servers to sense complex patterns in their market and environment, and make fast, accurate predictions that directly impact their bottom line". MapR complements RAPIDS with a data management and logistics fabric to accelerate the high-scale processing and access of disparate data across geographies. Oracle is also working to support the platform on its Oracle Data Science Cloud, it said. "These technologies are driving RAPIDS' ability to integrate into today's most popular data science workloads and accelerate them.... Artificial intelligence, analytics and machine learning technology can play a critical role in uncovering insights that can help customers achieve breakthrough results and improve the world we live in", said Antonio Neri, CEO, Hewlett Packard Enterprise.

"We are very excited about this collaboration with NVIDIA". The companies are also announcing the general availability of support for GPU-accelerated deep learning and high performance computing (HPC) containers from the (NGC) container registry on Oracle Cloud Infrastructure. "RAPIDS is accelerating the speed at which this processing and machine learning training can be done", said Clay Magouyrk, senior vice president of Software Development, Oracle Cloud Infrastructure.

Access to the RAPIDS open-source suite of libraries is immediately available online, where the code is being released under the Apache license. With this new offering and support for NGC containers, Oracle and NVIDIA are allowing customers to deploy containerized applications and frameworks for HPC, data science, and AI and run them on Oracle Cloud Infrastructure.

Other reports by Iphone Fresh

Discuss This Article