Amdocs and NVIDIA to Accelerate Adoption of Generative AI for $1.7 Trillion Telecom Industry

Amdocs and NVIDIA today announced they are collaborating to optimize large language models (LLMs) to speed adoption of generative AI applications and services across the $1.7 trillion telecommunications and media industries.(1)

Amdocs and NVIDIA will customize enterprise-grade LLMs running on NVIDIA accelerated computing as part of the Amdocs amAIz framework. The collaboration will empower communications service providers to efficiently deploy generative AI use cases across their businesses, from customer experiences to network provisioning.

Amdocs will use NVIDIA DGX Cloud AI supercomputing and NVIDIA AI Enterprise software to support flexible adoption strategies and help ensure service providers can simply and safely use generative AI applications.

Aligned with the Amdocs strategy of advancing generative AI use cases across the industry, the collaboration with NVIDIA builds on the previously announced Amdocs-Microsoft partnership. Service providers and media companies can adopt these applications in secure and trusted environments, including on premises and in the cloud.

With these new capabilities — including the NVIDIA NeMo framework for custom LLM development and guardrail features — service providers can benefit from enhanced performance, optimized resource utilization and flexible scalability to support emerging and future needs.

“NVIDIA and Amdocs are partnering to bring a unique platform and unmatched value proposition to customers,” said Shuky Sheffer, Amdocs Management Limited president and CEO. “By combining NVIDIA’s cutting-edge AI infrastructure, software and ecosystem and Amdocs’ industry-first amAlz AI framework, we believe that we have an unmatched offering that is both future-ready and value-additive for our customers.”

“Across a broad range of industries, enterprises are looking for the fastest, safest path to apply generative AI to boost productivity,” said Jensen Huang, founder and CEO of NVIDIA. “Our collaboration with Amdocs will help telco service providers automate personalized assistants, service ticket routing and other use cases for their billions of customers, and help the telcos analyze and optimize their operations.”

Amdocs counts more than 350 of the world’s leading telecom and media companies as customers, including 27 of the world’s top 30 service providers.(2) With more than 1.7 billion daily digital journeys, Amdocs platforms impact more than 3 billion people around the world.

NVIDIA and Amdocs are exploring a number of generative AI use cases to simplify and improve operations by providing secure, cost-effective and high-performance generative AI capabilities.

Initial use cases span customer care, including accelerating customer inquiry resolution by drawing information from across company data. On the network operations side, the companies are exploring how to proactively generate solutions that aid configuration, coverage or performance issues as they arise.

(1) Source: IDC, OMDIA, Factset analyses of Telecom 2022-2023 revenue.
(2) Source: OMDIA 2022 revenue estimates, excludes China.

Editor’s Note:

Generative AI uses a variety of AI models, including: 

  • Language models: These models, like OpenAI’s GPT-3, generate human-like text. One of the most popular examples of language-based generative models are called large language models (LLMs).
  • Large language models are being leveraged for a wide variety of tasks, including essay generation, code development, translation, and even understanding genetic sequences.
  • Generative adversarial networks (GANs): These models use two neural networks, a generator, and a discriminator.
  • Unimodal models: These models only accept one data input format.
  • Multimodal models: These models accept multiple types of inputs and prompts. For example, GPT-4 can accept both text and images as inputs.
  • Variational autoencoders (VAEs): These deep learning architectures are frequently used to build generative AI models.
  • Foundation models: These models generate output from one or more inputs (prompts) in the form of human language instructions.
Other types of generative AI models include:  Neural networks, Genetic algorithms, Rule-based systems, Transformers, LaMDA, LLaMA, BLOOM, BERT, RoBERTa. 

Cloud Service Providers struggle with Generative AI; Users face vendor lock-in; “The hype is here, the revenue is not”

Global Telco AI Alliance to progress generative AI for telcos

Bain & Co, McKinsey & Co, AWS suggest how telcos can use and adapt Generative AI

Generative AI Unicorns Rule the Startup Roost; OpenAI in the Spotlight

Generative AI in telecom; ChatGPT as a manager? ChatGPT vs Google Search

Generative AI could put telecom jobs in jeopardy; compelling AI in telecom use cases


Intentional or Accident: Russian fiber optic cable cut (1 of 3) by Chinese container ship under Baltic Sea

From Reuters:

A Russian fiber optic cable under the Baltic Sea was completely severed last month when a Chinese container ship passed over it, state company Rostelecom said on Tuesday.

Finnish investigators have already said they suspect the vessel, the NewNew Polar Bear, of causing serious damage to the nearby Balticconnector gas pipeline by dragging its anchor over the sea bed during the same voyage.

Two other Baltic telecoms cables were damaged on the same night of October 7th, along the route that the ship was travelling, according to shipping data reviewed by Reuters.

The incidents have highlighted the vulnerability of marine cables and pipelines at a time when security fears are running high because of the Ukraine war. Investigators have yet to establish who was responsible for blowing up Russia’s Nord Stream gas pipelines under the Baltic last year.

A Rostelecom spokesperson, responding to emailed questions from Reuters, said the double armored fiber optic cable, with a thickness of 40.4 mm (1.6 inches), had been cut completely.

Asked if the company believed the Chinese ship had caused the damage, the spokesperson said: “At the time of the damage to the fiber optic cable, the Chinese ship New Polar Bear was at a point with coordinates coinciding with the route of the communication line.”

China has said it is willing to provide necessary information on the incident in accordance with international law. NewNew Shipping, the owner and operator of the NewNew Polar Bear, has previously declined to comment when contacted by Reuters.

In a statement earlier on Tuesday, Rostelecom publicly acknowledged the damage to its cable for the first time, describing it as an accident and without mentioning the cause.  It said the site of the damage was only 28 km (17 miles) from where the Balticconnector gas pipeline was ruptured soon afterwards.

In total, three Baltic telecoms cables and one pipeline were damaged in the space of less than nine hours.

Data from shipping intelligence firm MarineTraffic, reviewed by Reuters, showed that the New Polar Bear passed over a Swedish-Estonian telecoms cable at 1513 GMT, then over the Russian cable at around 2020 GMT, the Balticconnector at 2220 GMT and a Finland-Estonia telecoms line at 2349 GMT.

Rostelecom said the damage to its cable was recorded at 2030 GMT.

As far back as Oct. 13, President Vladimir Putin dismissed as “complete rubbish” suggestions that Russia might have been to blame for the Balticconnector damage and floated the possibility that a ship’s anchor could have caused it.

On Tuesday, the Kremlin referred further questions to the Communications Ministry, which did not respond to a Reuters request for comment.

Finnish police announced on Oct. 24 that they had found a ship’s anchor near the broken gas pipeline. They have not concluded whether the damage was caused accidentally or deliberately.  Operator Gasgrid has said the pipeline could be out of commission until April or longer.

Rostelecom said a specialised vessel had started repairs on the fiber optic cable on Sunday and that the work was expected to take 10 days, depending on weather conditions.

The cable runs from St Petersburg to Russia’s Baltic exclave of Kaliningrad. The company said users had not been affected because data was transmitted via terrestrial routes and backup satellite channels.


China seeks to control Asian subsea cable systems; SJC2 delayed, Apricot and Echo avoid South China Sea

Sabotage or Accident: Was Russia or a fishing trawler responsible for Shetland Island cable cut?

Geopolitical tensions arise in Asia over subsea fiber optic cable projects; U.S. intervened to flip SeaMeWe-6 contractor


Verizon once again delays 5G Standalone (SA) commercial service

Like AT&T, Verizon has promised 5G standalone (SA) core network for a very long time.  The mostly wireless U.S. carrier initially said it would launch standalone 5G in 2020. Some in the industry thought it did so in 2022. But the company said the technology ‘is in testing now’ and is still not available commercially.

“We have it in trials only at this point. We don’t have it commercially available for our customers,” Verizon’s chief networking executive, Joe Russo, said on a podcast last month hosted by Recon Analytics. “So more to come in the next several months as Verizon will be entering the standalone core game.”

“It is absolutely a capability that we think will be another enabler to new use cases. But … the reliability and performance of Verizon’s network is what we stand for, and I don’t put technology out into the network that is a step back. It has to be a step forward. And all of the data that I see – both internal testing and with external testing that happens out there in the market – tells me that SA [standalone] needs a little bit more time.”

“We’re doing significant developing and testing to make sure that both the data session and the voice sessions in a standalone world are as good or better than what you would expect in our 4G network today. So we see that in the next several months we’re going to get there, but it was not my goal to be first in deploying standalone. It’s my goal to be best in deploying standalone.”


Verizon spokesperson Kevin King clarified that “we have commercial traffic running on our 5G non standalone core. That is what we announced earlier in the year. Joe was referring to our 5G standalone core which is in testing now.”

That cop-out was contradicted by a statement made during a webinar for analysts on September 29th, which was obtained by Light Reading.  “People talk about the standalone core. Just terminology-wise, that’s the 5G core essentially. If you guys have read the stuff we’ve said publicly, certainly we serve some customers on portions of our 5G core,” said Mike Haberman, Verizon’s SVP of strategy and transformation, And then we have some internal stuff going on with other functionality on the core. We’re in the process of rolling out (5G SA) in a very smart fashion.”

“Here’s the deal: When you go to the standalone core, you can’t aggregate your LTE carriers. With the non standalone core I’m aggregating together both 5G and 4G. So when you go standalone you start to bifurcate the spectrum. So that’s the impact to the RAN [radio access network]. So you better be sure that your mobile [customer] distribution, where they are geography, makes sense. Or what will happen is those customers will experience a lower service level. No good. We want to be careful of that. So that’s why, when you do the standalone core, you have to pay very close attention to your radio access network because they are directly attached.”

On April 27th Verizon issued a press release describing the benefits of 5G standalone (SA) technology and how it’s “what sets Verizon apart.” However, the release doesn’t specifically say that Verizon launched the technology.  That despite Verizon last year announced it had begun moving traffic onto its new 5G core, which supports both the non standalone (NSA) and standalone (SA) versions of the technology.

Last year, Mobile World Live reported that Verizon was migrating “commercial traffic onto SA 5G core.” The article cited an unnamed Verizon representative. Mobile World Live also reported that Ericsson, Casa Systems, Oracle and Nokia supply Verizon’s 5G core.

Dell’Oro Group, in January 2023, listed Verizon among the few North American wireless providers that had commercially launched the technology.

“This is a moving target,” Recon Analytics analyst Roger Entner told Light Reading. But Entner said Verizon’s position on the standalone version of 5G makes sense. “The benefits you can get today from standalone are limited.”

–>This author totally disagrees with Mr. Entner, because TRUE 5G=5G SA.  IN OTHER WORDS, ALL OF THE 3GPP DEFINED 5G FEATURES REQUIRE 5G SA!  That includes 5G security and network slicing.


Light Reading’s Mike Dano wrote:

Verizon now appears to be roughly three years behind its initial standalone 5G rollout plans. In the summer of 2020, Verizon said it would begin moving traffic onto its standalone 5G core “in the second half of 2020 with full commercialization in 2021.”

Then, in early 2022, Verizon CTO Kyle Malady suggested that the operator would begin moving some of its fixed wireless access (FWA) traffic onto its standalone 5G core by June of that year. He also said at the time that Verizon would start putting smartphone traffic onto that core in 2023.


T-Mobile US and Dish Wireless are the only two 5G carriers that have launched commercial 5G SA.   AT&T has made a lot of noise about it’s 5G SA plans but has yet to launch.

AT&T’s chief networking executive, Chris Sambar, wrote in a September 29th blog post that AT&T was moving some customers to standalone 5G. “Many of the newest mobile devices are ready for 5G standalone, and we continue to move thousands of customers every day. We also recently launched AT&T Internet Air home fixed wireless service, and from the start, this product rides on standalone 5G.”


AT&T touts 5G advances; will deploy Standalone 5G when “the ecosystem is ready”- when will that be?

Analysys Mason: 40 operational 5G SA networks worldwide; Sub-Sahara Africa dominates new launches

GSA 5G SA Core Network Update Report

5G subscription prices rise in U.S. without killer applications or 5G features (which require a 5G SA core network)


SpaceX has majority of all satellites in orbit; Starlink achieves cash-flow breakeven

SpaceX accounts for roughly one-half of all orbital space launches around the world, and it’s growing its launch frequency. It also has a majority of all the satellites in orbit around the planet.  This Thursday, majority owner & CEO Elon Musk tweeted, “Excited to announce that SpaceX Starlink has achieved breakeven cash flow! Starlink (a SpaceX subsidiary) is also now a majority of all active satellites and will have launched a majority of all satellites cumulatively from Earth by next year.”

There are some 5,000 Starlink satellites in orbit. Starlink satellites are small, lower-cost satellites built by SpaceX that deliver high-speed, space-based internet service to customers on Earth. Starlink can cost about $120 a month and there is some hardware to buy as well.

Starlink ended 2022 with roughly 1 million subscribers. The subscriber count now isn’t known, but it could be approaching 2 million users based on prior growth rates. SpaceX didn’t return a request for comment.

In 2021, Musk said SpaceX would spin off and take Starlink public once its cash flow was reasonably predictable.

A SpaceX rocket carriers Starlink satellites into orbit. PHOTO CREDIT:  SPACEX

Starlink has been in the spotlight since last year as it helps provide Ukraine with satellite communications key to its war efforts against Russia.

Last month, Musk said Starlink will support communication links in Gaza with “internationally recognized aid organizations” after a telephone and internet blackout isolated people in the Gaza Strip from the world and from each other.

Musk has sought to establish the Starlink business unit as a crucial source of revenue to fund SpaceX’s more capital-intensive projects such as its next-generation Starship, a giant reusable rocket the company intends to fly to the moon for NASA within the next decade.

Starlink posted a more than six-fold surge in revenue last year to $1.4 billion, but fell short of targets set by Musk, the Wall Street Journal reported in September, citing documents.

SpaceX is valued at about $150 billion and is one of the most valuable private companies in the world.


Verizon transports 1.2 terabytes per second of data across a single wavelength

Verizon has upgraded its optical to electrical conversion cards to send data at speeds of 1.2 Tbps on a single wavelength through the carrier’s live production network. The trials demonstrated increased reliability and overall capacity as well, Verizon said.

The trials, which were conducted in metro Long Island, N.Y., were in partnership with Cisco and included technology from Acacia, as well. They utilized Cisco’s NCS 1014 transceiver shelf and Acacia’s Coherent Interconnect Module (CIM 8). Verizon said the module features silicon semiconductor chips with 5nm complementary metal-oxide semiconductor (CMOS) digital processing and 140 Gbaud silicon photonics using 3D packaging technology. In short, digital processing capabilities and transistor density both are increased.

Verizon said that it transmitted a 1.0 Tbps single wavelength through the Cisco NCS 20000 line system for more than 205 km. It traversed 14 fiber central offices (COs). The carrier said this is significant because progressive filtering and signal-to-noise degradation impact wavelengths as they pass through each CO. The trials also featured 800 Gbps transmission for 305 km through 20 COs — and a 1.2 Tbps wavelength that traversed three offices.

“We have bet big on fiber. Not only does it provide an award-winning broadband experience for consumers and enterprises, it also serves as the backbone of our wireless network. As we continue to see customers using more data in more varied ways, it is critical we continue to stay ahead of our customers’ demands by using the resources we have most efficiently,” said Adam Koeppe, SVP of Technology Planning at Verizon.

Image courtesy of Verizon

In addition to increasing data rates, the new optics technology from Cisco reduces the need for regeneration of the light signal (conversion to electrical and back to optical signals) along the path by compensating for the degradation of the light signal traveling through the fiber cable. This adds reliability and leads to a reduced cost per bit operating expense for more efficient network management.

Bill Gartner, senior vice president and general manager of Cisco Optical Systems and Optics, added, “This trial demonstrates our commitment to continuous innovation aimed at increasing wavelength capacity and reducing costs. The Verizon infrastructure built with the Cisco NCS 2000 open line system supports multiple generations of optics, thus protecting investments as technology evolves.”

In March, Windstream Wholesale said that it sent a 1 Tbps wave across its Converged Optical Network (ICON) between Dallas and Tulsa, a distance of 541 km.


Verizon Touts 1.2 Tbps Wavelengths Over Production Network – Telecompetitor

AT&T, Verizon and Comcast all lost fixed broadband subscribers in 2Q-2023

Page 3 of 3
1 2 3