Highlights:
- IBM is introducing new compute options that provide access to graphics devices for its cloud platform.
- The new Watsonx product suite from IBM consists of three distinct offerings. Each concentrates on simplifying a distinct aspect of AI initiatives for businesses.
IBM Corporation introduced Watsonx, a product suite designed to help businesses facilitate creating and deploying artificial intelligence models.
Watsonx was previewed alongside several other new offerings at the company’s annual Think conference. IBM is introducing new compute options that provide access to graphics cards for its cloud platform. In addition, the company will release a tool enabled by artificial intelligence to help customers monitor the carbon emissions due to their cloud usage.
About Watsonx
The new Watsonx product suite from IBM consists of three distinct offerings. Each concentrates on simplifying a distinct aspect of AI initiatives for businesses.
Watsonx.ai, the company’s first product, is a software infrastructure that can be used to train generative AI models and other kinds of neural networks. According to IBM, clients will not be required to construct their models from zero. Watsonx.ai will provide access to a catalog of preconfigured AI models and training datasets.
Fm.code is a set of built-in AI models designed to increase the productivity of software development teams. It permits programmers to generate code using natural language commands. Customers will also have access to fm.NLP, which is a compilation of LLMs optimized for industry-specific duties, according to IBM.
fm.geospatial is the name of a third neural network compilation being released concurrently. The models in the package are trained using climate and “remote sensing” data. IBM claims businesses can use fm.geospatial to monitor biodiversity and natural disaster patterns.
In addition to IBM-developed models, Watsonx.ai will provide access to Hugging Face’s open-source neural networks. Hugging Face, operated by the same-named startup, is a platform similar to GitHub, used for hosting open-source AI models and related files. IBM asserts that clients can access thousands of neural networks and datasets via the platform.
watsonx.data is the second component of IBM’s new Watsonx product suite. According to the company, it is a data repository where organizations can retain the data used to train their AI models.
When an artificial intelligence model requests data from a database, each of the requests is processed by a query engine. It is responsible for guaranteeing the most efficient file retrieval feasible. Different query engines retrieve data in a variety of methods.
watsonx.data will support several query engines to assist customers in optimizing their AI applications. Each application can utilize the information retrieval module that best suits its needs. IBM claims that watsonx.data will also offer several automation features to facilitate user work.
The third product suite component is termed watsonx.governance. It can detect both AI bias and model drift, technical issue that causes neural networks’ accuracy to degrade over time. IBM claims that watsonx.governance will also enable businesses to ensure AI projects adhere to privacy regulations.
Arvind Krishna, Chief Executive Officer, said, “Foundation models make deploying AI significantly more scalable, affordable, and efficient. We built IBM watsonx for the needs of enterprises so that clients can be more than just users; they can become AI advantaged. With IBM watsonx, clients can quickly train and deploy custom AI capabilities across their entire business, all while retaining full control of their data.”
New AI Offerings
In July, the fundamental components of the Watsonx suite will be made generally available. They will be joined by several additional offerings that IBM detailed recently at its Think event.
IBM is revising its public cloud platform, IBM Cloud, with new compute options providing access to Nvidia Corp graphics processing units. The company says GPUs can be used for AI training and inference duties. Inference is the process of deploying AI models in production following training.
The IBM Cloud Carbon Calculator will accompany the introduction of the new GPU options. According to the firm, machine learning is used to assist businesses in measuring the environmental impact of their technology infrastructure. The tool will become accessible to the public later this year.
In parallel, IBM intends to incorporate foundation models into multiple additional products. Watson Code Assistant is a tool that can generate code based on natural language queries. IBM will release AIOps Insights, an application meant to expedite troubleshooting tasks for infrastructure teams
IBM is introducing novel AI capabilities for business users. According to the company, the fm.geospatial model from the new Watsonx product portfolio will be integrated into its Environmental Intelligence Suite for analyzing climate and geospatial data. Watson Assistant and Watson Orchestrate, two IBM products built to automate common business tasks such as customer support ticket processing, will implement a distinct foundation paradigm.