After being away from the HPC systems business for a long time, IBM today launches its genAI and data platform, IBM watsonx, to establish itself in the HPC-AI and generative AI market. Announced at its annual Think conference in San Francisco, IBM has announced GPU-as-a-service infrastructure for AI-intensive workloads, an AI-powered dashboard that reports cloud carbon footprint, and We also announced the watsonx and genAI practices by IBM Consulting. AI Deployment for Clients.
According to IBM, watsonx is an AI development studio that includes IBM’s curated and trained foundational and open-source models, a data store that enables collection and cleansing of training and tuning data, and an AI development studio that provides purpose to the business. Get access to a toolkit for governance. An end-to-end AI workflow designed to make it easy to adapt and scale AI.
“With the development of our foundational model, AI for business is stronger than ever. Arvind Krishna, Chairman and CEO of IBM. “The underlying model makes AI deployments much more scalable, affordable, and efficient. With IBM watsonx, our customers can rapidly train and deploy custom AI capabilities across their business while maintaining complete control over their data.”
According to IBM, the new offering is built for companies looking to AI, needing access to a technology stack for deploying AI models, including foundational models and machine learning capabilities, and relying on trusted data. , speed, and governance to run in any cloud environment.
The IBM watsonx platform consists of three product sets:
IBM watsonx.ai: Enterprise Studio Coming Generally Available July 2023enabling AI builders to deploy both traditional machine learning and new generative AI capabilities powered by underlying models through an open and intuitive user interface.
- AI Studio provides foundational models, training and tuning tools, and infrastructure that facilitate the entire data and AI lifecycle, from data preparation to model development, deployment, and monitoring.
- It also includes a library of IBM foundation models, using curated sets of enterprise data with filtering and cleansing processes and auditable data lineage. These models are trained on different modalities such as code, time series data, tabular data, geospatial data, IT event data, as well as language. The model categories are:
- fm.code: A model built to automatically generate code for developers via a natural language interface, increasing developer productivity and enabling automation of many IT tasks.
- fm.NLP: A collection of Large Language Models (LLMs) for a specific or industry-specific domain, leveraging curated data that can more easily mitigate bias and be rapidly customized with client data To do.
- fm.geospatial: A model built on climate and remote sensing data that allows organizations to understand natural disaster patterns, biodiversity, land use, and other geophysical processes that can affect their business. It helps you understand and plan for change.
- The watsonx.ai studio is built on Hugging Face’s open source library and provides thousands of Hugging Face open models and datasets. This is part of IBM’s commitment to providing clients with an open ecosystem approach that allows them to leverage the models and architectures that best suit their unique business needs.
IBM watsonx.data: A purpose-built data store built on an open lakehouse architecture. It’s optimized for governed data and AI workloads, supported by query, governance, and open data formats for accessing and sharing data. This solution is July 2023 and:
- This solution can manage workloads in both on-premises and multi-cloud environments.
- Through workload optimization, this solution can help organizations reduce data warehouse costs by up to 50%.1
- Watsonx.data enables users to access increasingly robust data from a single entry point, applying multiple fit-for-purpose query engines to uncover valuable insights. can.
- It also offers built-in governance tools, automation, and integration with your organization’s existing databases and tools to simplify setup and user experience.
IBM watsonx.governance: AI governance toolkit that enables trustworthy AI workflows. This solution will be generally available later this year.
- We operationalize governance to reduce the risk, time and costs associated with manual processes and provide the documentation you need to drive transparent and accountable results.
- It provides mechanisms to protect customer privacy, proactively detect model bias and drift, and help organizations meet ethical standards.
The watsonx platform enables clients to meet organizational needs in five key areas of business. Automation of business workflows and internal processes. Automation of IT processes. Threat protection. Commit to sustainability goals.
IBM also plans to inject the watsonx.ai foundational model into all major future software products. For example:
- Watson Code Assistant: A solution that uses generative AI to help developers generate code with simple English commands, coming later this year.
- AIOps Insights: AI Operations (AIOps) capabilities are enhanced with underlying models expected from code and NLP to provide greater visibility into performance across IT environments, enabling IT Operations (ITOps) Managers and Site Reliability Engineers (SREs) to ) to resolve incidents more quickly and cost-effectively. – Efficient method.
- Watson Assistant and Watson Orchestrate: IBM’s digital labor products, combined with NLP foundational models, are expected to improve employee productivity and customer service experiences.
- Environmental Intelligence Suite: IBM EIS Builder Edition is enabled by geospatial foundation models and will be available in preview later this year. It enables organizations to create customized solutions to address and mitigate environmental risks based on their unique goals and needs.
Also at Think 2023, IBM will announce additional upcoming offerings planned to accelerate the adoption of AI, including:
- New GPU offerings on IBM Cloud: To meet the global need for foundational models, IBM is announcing new GPU offerings on IBM Cloud. IBM Cloud is an AI-tuned infrastructure designed to support enterprise compute-intensive workloads. Later this year, IBM plans to offer a full-stack, high-performance, flexible AI-optimized infrastructure as a service on the IBM Cloud for both training and service foundation models.
- IBM Consulting Center of Excellence for Generative AI: IBM Consulting Announces Center of Excellence for Generative AI with Over 1,000 Generative AI Experts Focused on Watsonx, Actively Building and Deploying Watsonx for Customers We plan to build a practice that focuses on IBM Consulting has completed his engagements with dozens of clients by infusing generative AI into IBM Watson and a portfolio of ecosystem partners through the proven IBM Garage method.
- IBM Cloud Carbon Calculator: An AI-powered dashboard that enables clients to measure, track, manage and report on the carbon footprint associated with their hybrid multicloud journey. The IBM Cloud Carbon Calculator, based on technology from IBM Research, will be generally available later this year. The dashboard complements IBM’s existing sustainability solutions with a comprehensive portfolio of technology and expertise, including IBM Envizi ESG Suite, IBM Turbonomic, IBM Planning Analytics and IBM LinuxONE, to help organizations integrate sustainability and business Helps accelerate goals.
- New Research Reveals Generative AI Is One of Seven Trends Shaping Business: A new report from the IBM Institute for Business Value, Seven Bets, shares seven trends impacting businesses today, Learn about 7 bets worth making to strengthen your business. We need to adopt an “AI first” mindset and adopt ways in which leaders can most effectively leverage current and future AI opportunities and manage increased risk across the organization.
The Think 2023 broadcast can be accessed at www.ibm.com/events/think.keynote speech Tuesday, May 9, 8:30am – 9:15am EDTIBM CEO and Chairman Arvind Krishna Leaders from Delta Airlines, Citi, and Red Hat will join us on stage to discuss our vision of using hybrid cloud and AI to creatively turn challenges into opportunities.upon Wednesday, May 10, 4:30pm – 5pm EDTAt , IBM Research Senior Vice President and Director Dario Gill will deliver a keynote on how generative AI can be applied to business use cases.
Following the flagship event at OrlandoThink on Tour travels to more than 10 cities around the world: https://www.ibm.com/events/think/on-tour/.
