Recently, IBM revealed its plans to integrate the new open-source software RAPIDS into its enterprise-based data science platform designed for multi-cloud, hybrid and on-premises environments.
Looking at IBM ’s extensive portfolio made up of machine learning and deep learning solutions, the company appears to be well-positioned to deliver this open-source technology, specifically to data scientists irrespective of their chosen deployment model.
“IBM has a long collaboration with NVIDIA that has shown demonstrable performance increases leveraging IBM technology, like the IBM POWER9 processor, in combination with NVIDIA GPUs. We look to continue to aggressively push the performance boundaries of AI for our clients as we bring RAPIDS into the IBM portfolio,” explained Bob Picciano, the Senior Vice President of IBM Cognitive Systems.
MORE – IBM Watson: What you Need to Know
RAPIDS is expected to deliver GPU acceleration features to IBM offerings, which leverage open source machine learning software such as scikit-learn, Pandas and Apache Arrow. The immediate, extensive ecosystem support for the new open-source software RAPIDS stems from various core open-source contributors such as Ursa Labs, INRIA, PyData, NERSC, Graphistry, Anaconda, and BlazingDB.
IBM intends to deliver RAPIDS to various such as multi-cloud, hybrid, public and on-premises environments such as:
- PowerAI on IBMPOWER9, to take advantage of RAPIDS in expanding the available options to data scientists equipped with new open-source analytics and machine learning libraries. PowerAI prides itself on being IBM’s software layer that improves how today’s AI workloads and data science operate on heterogeneous computing systems.
- IBM Cloud, to users who select machines fitted with GPUs, will be in a better position to apply the analytics libraries in RAPIDS and accelerated machine learning not only for their cloud applications but also tapping the benefits of machine learning.
- IBM Watson Studio and Watson Machine Learning, to leverage the capability of NVIDIA GPUs in a bid to allow AI developers and data scientists to create, deploy and operate faster models compared to the CPU-only deployments designed for their AI applications, particularly in a multi-cloud environment featuring IBM Cloud and IBM Cloud Private.
“IBM and NVIDIA’s close collaboration over the years has helped leading enterprises and organizations around the world tackle some of the world’s largest problems. Now, with IBM taking advantage of RAPIDS open-source libraries announced today by NVIDIA, GPU accelerated machine learning is coming to data scientists, helping them analyze big data for insights faster than ever possible before, ” asserted Ian Buck, VP and general manager of Accelerated Computing at NVIDIA.
Machine learning acts as a form of AI that allows a system to learn from data as opposed to explicit programming. Businesses across various industries such as telecommunications, finance, and retail are either utilizing machine learning or exploring machine learning thanks to the potential value it provides to companies that are striving to take advantage of big data in a bid to assist them in comprehending the subtle transformations in customer satisfaction, preferences and behavior.
At the start of this year, IBM made headlines by setting a record, particularly in a tera-scale machine learning (ML) benchmark. It beat the company that held the previous record by 46 times.
