AT&T realizes huge value from AI; will use full suite of NVIDIA AI offerings

Executive Summary:

AT&T Corp. and NVIDIA today announced a collaboration in which AT&T will continue to transform its operations and enhance sustainability by using NVIDIA-powered AI for processing data, optimizing service-fleet routing and building digital avatars for employee support and training.

AT&T is the first telecommunications provider to explore the use of a full suite of NVIDIA AI offerings. This includes enhancing its data processing using the NVIDIA AI Enterprise software suite, which includes the NVIDIA RAPIDS Accelerator for Apache Spark; enabling real-time vehicle routing and optimization with NVIDIA cuOpt; adopting digital avatars with NVIDIA Omniverse Avatar Cloud Engine and NVIDIA Tokkio; and utilizing conversational AI with NVIDIA Riva.

“We strive each day to deliver the most efficient global network, as we drive towards net zero emissions in our operations,” said Andy Markus, chief data officer at AT&T. “Working with NVIDIA to drive AI solutions across our business will help enhance experiences for both our employees and customers.”  He said it’s AT&T’s goal to make AI part of the fabric of the company, to have “all parts of the business leveraging AI and creating AI” rather than limit its use to creation of AI by its specialist data scientists.

“Industries are embracing a new era in which chatbots, recommendation engines and accelerated libraries for data optimization help produce AI-driven innovations,” said Manuvir Das, vice president of Enterprise Computing at NVIDIA. “Our work with AT&T will help the company better mine its data to drive new services and solutions for the AI-powered telco.”

The Data Dilemma:
AT&T, which has pledged to be carbon neutral by 2035, has instituted broad initiatives to make its operations more efficient. A major challenge is optimizing energy consumption while providing network infrastructure that delivers data at high speeds.  AT&T processes more than 590 petabytes of data on average a day. That is the equivalent of about 6.5 million 4K movies or more than 8x the content housed in the U.S. Library of Congress if all its collections were digitized.

Telecoms aiming to reduce energy consumption face challenges across their operations. Within networks, the radio access network (RAN) consumes 73% of energy, while core network services, data centers and operations use 13%, 9% and 5%, respectively, according to the GSMA, a mobile industry trade group.

AT&T first adopted NVIDIA RAPIDS Accelerator for Apache Spark to capitalize on energy-efficient GPUs across its AI and data science pipelines. This helped boost its operational efficiency across everything from training AI models and maintaining network quality and optimization, to reducing customer churn and improving fraud detection.

Of the data and AI pipelines targeted with Spark-RAPIDS, AT&T saves about half of its cloud computing spend and sees faster performance, while enabling reductions in its carbon footprint.

Enhanced Field Dispatch Services:
AT&T, which operates one of the largest field dispatch teams to service its customers, is currently testing NVIDIA cuOpt software to enhance its field dispatch capabilities to handle more complex technician routing and optimization challenges.  AT&T has a fleet of roughly 30,000 vehicles with over 700 million options in how they can be dispatched and routed. The operator would run dispatch optimization algorithms overnight to get plans for the next day, but it took too long and couldn’t account for the realities that would crop up the next morning: Workers calling in sick, vehicles breaking down, and so on.

“It wasn’t as good at noon as it was at 8 in the morning,” Markus said. Using Nvidia GPUs and software, he said, AT&T was able to speed up its processing 60x so that it could run the scenario in near-real-time, as often as it needed to and achieve more jobs in a day (as well as reduce its cloud-related costs by 40%).

Routing requires trillions of computations to factor in a variety of factors, from traffic and weather conditions to customer change of plans or a technician’s skill level, where a complicated job might then require an additional truck roll.

In early trials, cuOpt delivered solutions in 10 seconds, while the same computation on x86 CPUs took 1,000 seconds. The results yielded a 40% reduction in cloud costs and allowed technicians to complete more service calls each day. NVIDIA cuOpt allows AT&T to run nearly continuous dispatch optimization software by combining NVIDIA RAPIDS with local search heuristics algorithms and metaheuristics such as Tabu search.

Pleasing Customers, Speeding Network Design:
As part of its efforts to improve productivity for its more than 150,000 employees, AT&T is moving to adopt NVIDIA Omniverse ACE and NVIDIA Tokkio, cloud-native AI microservices, workflows and application frameworks for developers to easily build, customize and deploy interactive avatars that see, perceive, intelligently converse and provide recommendations to enhance the customer service experience.

For conversational AI, the carrier also uses the NVIDIA Riva software development kit and is examining other customer service and operations use cases for digital twins and generative AI.

AT&T also is taking advantage of fast 5G and its fiber network to deliver NVIDIA GeForce NOW™ cloud gaming at 120 frames per second on mobile and 240 FPS at home.

Markus added that AI-powered Nvidia tools are also helping AT&T to both serve its customers better through various channels, from sales recommendations to customer care; and that its internal processes are leveraging AI as well, to help employees be more efficient. The company is embracing Nvidia’s AI solutions as a foundation for development of interactive and intelligent customer service avatars.

In the past 12 months,  AI has created more than $2.5 billion in value for AT&T. About half of that came via Marcus’ team, but the other half came from what he calls “citizen data scientists” across the company who have been able to leverage AI to solve problems in their respective areas, whether than was marketing, network operations, software development or finance.

“As we mobilize that citizen data-scientist across the company, we’re doing that via a self-service platform that we call AI-as-a-service, where we’re bringing a unified experience together. But behind the experience, we’re allowing those users to leverage AI in a curated way for their use case,” he explained. “So they bring their subject matter expertise to the problem that they’re trying to solve, and we … enable the technology [and processes for them to create] robust AI. But we also govern it with some guardrails, so the AI we’re creating is ethical and responsible.”

In AT&T’s automation development, 92% of its automation is created by employees via self-service to solve a problem. “The goal is that over time, we bake in incredible functionality like Nvidia, so that AI-as-a-service is delivering that self-service functionality so that we do most of our routine AI creation via the platform, where you don’t have to have a professional data scientist, a code warrior, to be your sherpa,” Markus concluded.


AT&T leans into AI, and leans on Nvidia to do it

3 thoughts on “AT&T realizes huge value from AI; will use full suite of NVIDIA AI offerings

  1. Related: Medtronic and NVIDIA Collaborate to Build AI Platform for Medical Devices

    NVIDIA today announced that it is collaborating with Medtronic, the world’s largest healthcare technology provider, to accelerate the development of AI in the healthcare system and bring new AI-based solutions into patient care.

    The companies will integrate NVIDIA healthcare and edge AI technologies into Medtronic’s GI Genius™ intelligent endoscopy module, developed and manufactured by Cosmo Pharmaceuticals. GI Genius is the first FDA-cleared, AI-assisted colonoscopy tool to help physicians detect polyps that can lead to colorectal cancer.

    GI Genius has been designed to host a suite of AI algorithms and integrating the NVIDIA Clara™ healthcare platform could allow Medtronic to scale development of algorithms for real-time procedures, potentially accelerating AI innovation for better patient care.

    “Artificial intelligence is a powerful tool that can increase the speed, efficiency and effectiveness of global health systems,” said Kimberly Powell, vice president of healthcare at NVIDIA. “We’re collaborating with Medtronic to accelerate AI innovation by enabling a software-defined business model, with the goal of improving clinical decision making, reducing medical variability and driving better patient outcomes.

    “We believe that collaborating with AI companies and developers like NVIDIA and Cosmo is essential to driving innovation within the medical device industry,” said Giovanni Di Napoli, president of Gastrointestinal Business at Medtronic. “We are committed to working with the best and brightest minds in the field of AI to develop new technologies that can improve patient outcomes and transform the way we approach healthcare.”

    Medtronic intends to integrate NVIDIA Holoscan — a real-time AI computing software platform for building medical devices — and NVIDIA IGX, an industrial-grade edge AI hardware platform, to run with its GI Genius AI-assisted colonoscopy system to support physicians with AI-enhanced diagnostic images. Holoscan helps bring the latest AI applications into clinical settings by providing the full-stack infrastructure needed for scalable, software-defined processing of streaming data at the edge.

    The NVIDIA Holoscan and IGX platform makes software-defined medical devices possible by enabling developers to efficiently train and validate AI models within the Cosmo Innovation Center, and then host the AI-powered applications on Medtronic’s GI Genius AI Access™ Platform, a marketplace for software-as-a-medical-device (SaMD) applications. The first GI Genius systems built with the NVIDIA technology will be available later this year.

  2. Nvidia’s reach in artificial intelligence technology already spans graphics processing units, high-performance computing systems, software and cloud computing services. And Nvidia stock has been on a tear this year thanks to investor excitement over AI.

    The company on Tuesday revealed several partnerships and new products, demonstrating how its technology can advance artificial intelligence in myriad industries. Nvidia says ChatGPT, the AI program that’s created a firestorm of interest, is just the start. Nvidia issued about two dozen press releases and blog posts at its GTC 2023 conference outlining its AI plans.

    In a keynote presentation at the conference, Chief Executive Jensen Huang said artificial intelligence is at an inflection point.

    “We are at the iPhone moment of AI,” Huang said.
    At GTC, Huang announced partnerships with Alphabet’s (GOOGL) Google, Microsoft (MSFT), Oracle (ORCL) and others. The alliances aim to bring AI, simulation and collaboration capabilities to every industry, he said.

    “The impressive capabilities of generative AI have created a sense of urgency for companies to re-imagine their products and business models,” Huang said.

    At the company’s GTC conference, Nvidia announced new computer hardware and software libraries to enable AI breakthroughs.

    Nvidia also said it will offer its DGX AI supercomputer platform as a service through cloud-computing partners. The DGX Cloud service will run first on Oracle Cloud. Microsoft Azure, Google Cloud and others will follow soon, Huang said.

    DGX Cloud will give enterprises immediate access to the infrastructure and software needed to train advanced models for generative AI and other applications.

    Nvidia Touts Platforms For Generative AI
    Nvidia also announced Nvidia AI Foundations, a family of cloud services for customers needing to build and run custom large language models and generative AI trained with their proprietary data. Generative AI such as chatbot ChatGPT and text-to-image generator Dall-E have captured the imaginations of enterprises and investors.

    AI Foundations services come in three modalities. NeMo is for building text-to-text generative AI models. Picasso is for generating images, videos and 3D models from text. And BioNeMo is for the biotech industry to develop new drugs.

    Adobe (ADBE), Getty Images (GETY) and Shutterstock (SSTK) are using Picasso. Meanwhile, Amgen (AMGN) and AstraZeneca (AZN) are among the early BioNeMo customers.

    Also, Nvidia hopes to accelerate semiconductor breakthroughs with its cuLitho software library for computational lithography. The fabless semiconductor company is working on cuLitho with chip-gear vendor ASML (ASML), chip foundry TSMC (TSM) and chip design software firm Synopsys (SNPS).

    Huang said cuLitho will help achieve integrated circuits at 2 nanometers and smaller. Features on semiconductors are measured in nanometers, which are one-billionth of a meter.

  3. AT&T has narrowed its business to focus on mobile and broadband to achieve more stable and predictable growth, which is starting to yield results. The Mobility segment — its largest at over 65% of sales — has been posting steady service-revenue gains on postpaid phone net additions. This will likely continue as AT&T boosts network capacity and deploys mid-band spectrum to 160 million people, which is allowing it to more effectively compete in 5G. It’s also rapidly expanding the fiber network, which has yielded solid user and revenue increases. It plans to pass 30 million locations with fiber by year-end 2025.

Comments are closed.