Superclusters of Nvidia GPU/AI chips combined with end-to-end network platforms to create next generation data centers

Meta Platforms and Elon Musk’s xAI start-up are among companies building clusters of computer servers with as many as 100,000 of Nvidia’s most advanced GPU chips as the race for artificial-intelligence (AI) supremacy accelerates.

  • Meta Chief Executive Mark Zuckerberg said last month that his company was already training its most advanced AI models with a conglomeration of chips he called “bigger than anything I’ve seen reported for what others are doing.”
  • xAI built a supercomputer called Colossus—with 100,000 of Nvidia’s Hopper GPU/AI chips—in Memphis, TN in a matter of months.
  • OpenAI and Microsoft have been working to build up significant new computing facilities for AI. Google is building massive data centers to house chips that drive its AI strategy.

xAI built a supercomputer in Memphis that it calls Colossus, with 100,000 Nvidia AI chips. Photo: Karen Pulfer Focht/Reuters

A year ago, clusters of tens of thousands of GPU chips were seen as very large. OpenAI used around 10,000 of Nvidia’s chips to train the version of ChatGPT it launched in late 2022, UBS analysts estimate. Installing many GPUs in one location, linked together by superfast networking equipment and cables, has so far produced larger AI models at faster rates. But there are questions about whether ever-bigger super clusters will continue to translate into smarter chatbots and more convincing image-generation tools.

Nvidia Chief Executive Jensen Huang  said that while the biggest clusters for training for giant AI models now top out at around 100,000 of Nvidia’s current chips, “the next generation starts at around 100,000 Blackwells. And so that gives you a sense of where the industry is moving. Do we think that we need millions of GPUs? No doubt. That is a certainty now. And the question is how do we architect it from a data center perspective,” Huang added.

“There is no evidence that this will scale to a million chips and a $100 billion system, but there is the observation that they have scaled extremely well all the way from just dozens of chips to 100,000,” said Dylan Patel, the chief analyst at SemiAnalysis, a market research firm.

Giant super clusters are already getting built. Musk posted last month on his social-media platform X that his 100,000-chip Colossus super cluster was “soon to become” a 200,000-chip cluster in a single building. He also posted in June that the next step would probably be a 300,000-chip cluster of Nvidia’s newest GPU chips next summer.  The rise of super clusters comes as their operators prepare for Nvidia’s nexgen Blackwell chips, which are set to start shipping out in the next couple of months. Blackwell chips are estimated to cost around $30,000 each, meaning a cluster of 100,000 would cost $3 billion, not counting the price of the power-generation infrastructure and IT equipment around the chips.

Those dollar figures make building up super clusters with ever more chips something of a gamble, industry insiders say, given that it isn’t clear that they will improve AI models to a degree that justifies their cost. Indeed, new engineering challenges also often arise with larger clusters:

  • Meta researchers said in a July paper that a cluster of more than 16,000 of Nvidia’s GPUs suffered from unexpected failures of chips and other components routinely as the company trained an advanced version of its Llama model over 54 days.
  • Keeping Nvidia’s chips cool is a major challenge as clusters of power-hungry chips become packed more closely together, industry executives say, part of the reason there is a shift toward liquid cooling where refrigerant is piped directly to chips to keep them from overheating.
  • The sheer size of the super clusters requires a stepped-up level of management of those chips when they fail. Mark Adams, chief executive of Penguin Solutions, a company that helps set up and operate computing infrastructure, said elevated complexity in running large clusters of chips inevitably throws up problems.

The continuation of the AI boom for Nvidia largely depends on how the largest clusters of GPU chips deliver a return on investment for its customers. The trend also fosters demand for Nvidia’s networking equipment, which is fast becoming a significant business. Nvidia’s networking equipment revenue in 2024 was $3.13 billion, which was a 51.8% increase from the previous year.  Mostly from its Mellanox acquisition, Nvidia offers these networking platforms:

  • Accelerated Ethernet Switching for AI and the Cloud

  • Quantum InfiniBand for AI and Scientific Computing

  • Bluefield® Network Accelerators

………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………..

Nvidia forecasts total fiscal fourth-quarter sales of about $37.5bn, up 70%. That was above average analyst projections of $37.1bn, compiled by Bloomberg, but below some projections that were as high as $41bn. “Demand for Hopper and anticipation for Blackwell – in full production – are incredible as foundation model makers scale pretraining, post-training and inference, Huang said.  “Both Hopper and Blackwell systems have certain supply constraints, and the demand for Blackwell is expected to exceed supply for several quarters in fiscal 2026,” CFO Colette Kress said.

References:

https://www.wsj.com/tech/ai/nvidia-chips-ai-race-96d21d09?mod=tech_lead_pos5

https://www.datacenterdynamics.com/en/news/nvidias-data-center-revenue-up-112-over-last-year-as-ai-boom-continues/

https://www.nvidia.com/en-us/networking/

https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-third-quarter-fiscal-2025

HPE-Juniper combo + Cisco restructuring create enterprise network uncertainty

Hewlett Packard Enterprise’s (HPE) pending acquisition of Juniper Networks and Cisco’s recent corporate restructuring (which de-emphasizes legacy networking products like access/core routers and Ethernet switches) is putting enterprise networking customers in a holding pattern.   They are  pausing investments in network equipment as they wait out the uncertainty.

“I’ve had customers put things on hold right now, and not just the Juniper side but both sides,” Andre Kindness, principal analyst at Forrester Research, said in an interview with SDxCentral about how Juniper and HPE customers are reacting to uncertainty around the deal. “Typically, if customers are strong enough to look outside of Cisco and they’re not a Cisco shop, then HPE, Aruba, Juniper are the primary ones that they’re looking at. I’ve had customers put some of that on hold at this point.”

That holding pattern is tied to uncertainty over what systems and platforms will emerge from a combined HPE-Juniper. Mr. Kindness noted in a blog post when the deal was announced that “the journey ahead will be rife with obstacles for Juniper and HPE/Aruba customers alike.” Kindness explained that one important move for HPE would be to “rationalize/optimize the portfolio, the products and the solutions.”

“HPE will try to reassure you that nothing will change; it doesn’t make sense to keep everything, especially the multiple AP [access point] product lines (Instant On, Mist, and Aruba Aps), all the routing and switching operating systems (Juno, AOS-CX, and ArubaOS) and both management systems (Central and Mist),” Kindness wrote.

“Though not immediately, products will need to go and the hardware that stays will need to be changed to accommodate cloud-based management, monitoring, and AI.”  HPE CEO Antonio Neri and his management team has attempted to temper these concerns by stating there is virtually no overlap between HPE and Juniper’s product lines, which Kindness said, “just boggles my mind,” he added.

Juniper’s AI product, called Marvis (part of the Mist acquisition in 2019), is by far the most advanced AI solution in the networking market. That’s not a profound statement; no vendor has anything close to it. The quick history: Juniper’s acquisition of Mist brought the company a cloud-based Wi-Fi solution with a leading AI capability, Marvis. Juniper quickly started integrating its switching and routing portfolio into Marvis. Walmart, Amazon, and others took notice. Fast-forward to today: This gives HPE Aruba a two-year lead against its competitors by bringing Juniper into the fold.

“I think [Neri’s] got to worry about the financial analyst out there in the stock market or the shareholders to pacify them, and then at the same time you don’t want to scare the bejesus out of your customer base, or Juniper customer base, so you’re going to say that there’s going to be either no overlap or no changes, everything will coexist,” Kindness said.

While overlap and other concerns could alter what a potential Juniper HPE combo looks like, Kindness said he expects the result to lean heavily on Juniper’s telecom and networking assets. That includes HPE products like Aruba networking gear being replaced by Juniper’s artificial intelligence (AI)-focused Mist and Marvis platforms.

“Mist has been really a game changer for the company and just really opened a lot of doors,” Kindness explained. “[Juniper] really did a 180 degree turn when they bought [Mist], and just the revenue that’s brought in and the expansion of the product line itself, and the capabilities of Mist and actually Marvis in the background would be hard for [HPE] to replicate at this point. My perception was HPE looked at it and said, Marvis and Mist is just something that would take too long to get to.” Kindness added that he does not expect significant platform thinning to happen for a couple of years after a potential closing of the deal, but the interim could be filled with challenges tied to channel partners and go-to-market strategies that could chip away at market opportunities similar to what is happening at VMware following the Broadcom acquisition.  “Broadcom is ruthless, right or wrong, it’s its business model,” Kindness said. “HPE is not quite that dynamic.”

……………………………………………………………………………………………………………………………………….

Cisco CFO Scott Herren told the audience at a recent investor conference that HPE’s pending Juniper acquisition is causing “uncertainty” in the enterprise WLAN market that could be benefit Cisco. “I think for sure that’s created just a degree of uncertainty and a question of, hey, should I consider if I was previously a vendor or a customer of either of those, now is the time to kind of open up and look at other opportunities,” Herren said. “And we’ve seen our wireless business, our orders greater than $1 million grew more than 20% in the fourth quarter.”

Cisco is also working through its own networking drama as part of the vendor’s recently announced restructuring process. Those moves will see Cisco focus more on high-growth areas like AI, security, and cloud at the expense of its legacy operations, including the pairing down of its networking product lines.

“It looks like Cisco’s realizing that all the complexity of customer choice and all these variations and offering a zillion features is probably not the way to go. I think Chuck realized it,” Kindness said of Cisco’s efforts. “If you look at the ACI [Application Centric Infrastructure] and Cloud Dashboard for Nexus starting to consolidate, and then the Catalyst line and the Aironet line and the Meraki line are consolidating, it’s just the right move. The market has told them that for the last 10 years, it just took them a while to recognize it.”

References:

https://www.sdxcentral.com/articles/analysis/hpe-juniper-cisco-networking-chaos-has-enterprises-nervous/2024/11/

https://www.juniper.net/us/en.html

HPE + Juniper Networks Creates A Cisco Doppelganger

 

Cisco to lay off more than 4,000 as it shifts focus to AI and Cybersecurity

Cisco restructuring plan will result in ~4100 layoffs; focus on security and cloud based products

 

Canalys & Gartner: AI investments drive growth in cloud infrastructure spending

According to market research firm Canalys, global spending on cloud infrastructure services [1.] increased by 21% year on year, reaching US$82.0 billion in the 3rd quarter of 2024.  Customer investment in the hyperscalers’ AI offerings fueled growth, prompting leading cloud vendors to escalate their investments in AI. 

Note 1. Canalys defines cloud infrastructure services as services providing infrastructure (IaaS and bare metal) and platforms that are hosted by third-party providers and made available to users via the Internet. 

The rankings of the top three cloud service providers –  Amazon AWS, Microsoft Azure and Google Cloud – remained stable from the previous quarter, with these providers together accounting for 64% of total expenditure. Total combined spending with these three providers grew by 26% year on year, and all three reported sequential growth. Market leader AWS maintained a year-on-year growth rate of 19%, consistent with the previous quarter. That was outpaced by both Microsoft, with 33% growth, and Google Cloud, with 36% growth. In actual dollar terms, however, AWS outgrew both Microsoft and Google Cloud, increasing sales by almost US$4.4 billion on the previous year.

In Q3 2024, the cloud services market saw strong, steady growth. All three cloud hyperscalers reported positive returns on their AI investments, which have begun to contribute to their overall cloud business performance. These returns reflect a growing reliance on AI as a key driver for innovation and competitive advantage in the cloud.

With the increasing adoption of AI technologies, demand for high-performance computing and storage continues to rise, putting pressure on cloud providers to expand their infrastructure. In response, leading cloud providers are prioritizing large-scale investments in next-generation AI infrastructure. To mitigate the risks associated with under-investment – such as being unprepared for future demand or missing key opportunities – they have adopted over-investment strategies, ensuring their ability to scale offerings in line with the growing needs of their AI customers. Enterprises are convinced that AI will deliver an unprecedented boost in efficiency and productivity, so they are pouring money into hyperscalers’ AI solutions. Accordingly, cloud service provider capital spending (CAPEX) will sustain their rapid growth trajectories and are expected to continue on this path into 2025.

“Continued substantial expenditure will present new challenges, requiring cloud vendors to carefully balance their investments in AI with the cost discipline needed to fund these initiatives,” said Rachel Brindley, Senior Director at Canalys. “While companies should invest sufficiently in AI to capitalize on technological growth, they must also exercise caution to avoid overspending or inefficient resource allocation. Ensuring the sustainability of these investments over time will be vital to maintaining long-term financial health and competitive advantage.”

“On the other hand, the three leading cloud providers are also expediting the update and iteration of their AI foundational models, continuously expanding their associated product portfolios,” said Yi Zhang, Analyst at Canalys. “As these AI foundational models mature, cloud providers are focused on leveraging their enhanced capabilities to empower a broader range of core products and services. By integrating these advanced models into their existing offerings, they aim to enhance functionality, improve performance and increase user engagement across their platforms, thereby unlocking new revenue streams.”

Amazon Web Services (AWS) maintained its lead in the global cloud market in Q3 2024, capturing a 33% market share and achieving 19% year-on-year revenue growth. It continued to enhance and broaden its AI offerings by launching new models through Amazon Bedrock and SageMaker, including Anthropic’s upgraded Claude 3.5 Sonnet and Meta’s Llama 3.2. It reported a triple-digit year-on-year increase in AI-related revenue, outpacing its overall growth by more than three times. Over the past 18 months, AWS has introduced nearly twice as many machine learning and generative AI features as the combined offerings of the other leading cloud providers. In terms of capital expenditure, AWS announced plans to further increase investment, with projected spending of approximately US$75 billion in 2024. This investment will primarily be allocated to expanding technology infrastructure to meet the rising demand for AI services, underscoring AWS’ commitment to staying at the forefront of technological innovation and service capability.

Microsoft Azure remains the second-largest cloud provider, with a 20% market share and impressive annual growth of 33%. This growth was partly driven by AI services, which contributed approximately 12% to the overall increase. Over the past six months, use of Azure OpenAI has more than doubled, driven by increased adoption by both digital-native companies and established enterprises transitioning their applications from testing phases to full-scale production environments. To further enhance its offerings, Microsoft is expanding Azure AI by introducing industry-specific models, including advanced multimodal medical imaging models, aimed at providing tailored solutions for a broader customer base. Additionally, the company announced new cloud and AI infrastructure investments in Brazil, Italy, Mexico and Sweden to expand capacity in alignment with long-term demand forecasts.

Google Cloud, the third-largest provider, maintained a 10% market share, achieving robust year-on-year growth of 36%. It showed the strongest AI-driven revenue growth among the leading providers, with a clear acceleration compared with the previous quarter. As of September 2024, its revenue backlog increased to US$86.8 billion, up from US$78.8 billion in Q2, signaling continued momentum in the near term. Its enterprise AI platform, Vertex, has garnered substantial user adoption, with Gemini API calls increasing nearly 14-fold over the past six months. Google Cloud is actively seeking and developing new ways to apply AI tools across different scenarios and use cases. It introduced the GenAI Partner Companion, an AI-driven advisory tool designed to offer service partners personalized access to training resources, enhancing learning and supporting successful project execution. In Q3 2024, Google announced over US$7 billion in planned data center investments, with nearly US$6 billion allocated to projects within the United States.

Separate statistics from Gartner corroborate hyperscale CAPEX optimism.  Gartner predicts that worldwide end-user spending on public cloud services is on course to reach $723.4 billion next year, up from a projected $595.7 billion in 2024.  All segments of the cloud market – platform-as-a-service (PaaS), software-as-a-service (SaaS), desktop-as-a-service (DaaS), and infrastructure-as-a-service (IaaS) – are expected to achieve double-digit growth.

While SaaS will be the biggest single segment, accounting for $299.1 billion, IaaS will grow the fastest, jumping 24.8 percent to $211.9 million.

Gartner_cloud_forecast.jpg

Like Canalys, Gartner also singles out AI for special attention. “The use of AI technologies in IT and business operations is unabatedly accelerating the role of cloud computing in supporting business operations and outcomes,” said Sid Nag, vice president analyst at Gartner. “Cloud use cases continue to expand with increasing focus on distributed, hybrid, cloud-native, and multicloud environments supported by a cross-cloud framework, making the public cloud services market achieve a 21.5 percent growth in 2025.”

……………………………………………………………………………………………………………………………………………………………………………………………………..

References:

https://canalys.com/newsroom/global-cloud-services-q3-2024

https://www.telecoms.com/public-cloud/ai-hype-fuels-21-percent-jump-in-q3-cloud-spending

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

MTN Consulting: Top Telco Network Infrastructure (equipment) vendors + revenue growth changes favor cloud service providers

IDC: Public Cloud software at 2/3 of all enterprise applications revenue in 2026; SaaS is essential!

IDC: Cloud Infrastructure Spending +13.5% YoY in 4Q-2021 to $21.1 billion; Forecast CAGR of 12.6% from 2021-2026

IDC: Worldwide Public Cloud Services Revenues Grew 29% to $408.6 Billion in 2021 with Microsoft #1?

Synergy Research: Microsoft and Amazon (AWS) Dominate IT Vendor Revenue & Growth; Popularity of Multi-cloud in 2021

Google Cloud revenues up 54% YoY; Cloud native security is a top priority

WSJ: T-Mobile hacked by cyber-espionage group linked to Chinese Intelligence agency

According to the Wall Street Journal, T-Mobile’s network was hacked in a damaging Chinese cyber-espionage operation that successfully gained entry into multiple U.S. and international telecommunications companies.

Hackers linked to a Chinese intelligence agency were able to breach T-Mobile as part of monthslong campaign to spy on the cellphone communications of high-value intelligence targets. It is unclear what information, if any, was taken about T-Mobile customers’ calls and communications records.

“T-Mobile is closely monitoring this industry-wide attack, and at this time, T-Mobile systems and data have not been impacted in any significant way, and we have no evidence of impacts to customer information,” a company spokeswoman said. “We will continue to monitor this closely, working with industry peers and the relevant authorities.”

The compromise of T-Mobile expands the list of known victims of a cyber-espionage campaign by Chinese hackers—dubbed Salt Typhoon—that some U.S. officials consider to be historic and catastrophic in scope and severity. The WSJ had reported in October that AT&T, Verizon and Lumen Technologies were among the telecom companies that suffered an intrusion.  The widespread compromise is considered a potentially catastrophic security breach. It appeared to be geared toward intelligence collection, people familiar with the matter said.

China’s multipronged spying operations have drawn warnings in the U.S. about their economic implications. Photo: Andy Wong/Associated Press

…………………………………………………………………………………………………………………………………………………………………………..

Salt Typhoon used sophisticated methods to infiltrate American telecom infrastructure through vulnerabilities including Cisco Systems routers, and investigators suspect the hackers relied on artificial intelligence or machine learning to further their espionage operations , people familiar with the matter said. The attackers penetrated at least some of that infrastructure over eight months or more.

In the broader hacking campaign, attackers were able to access cellphone lines used by an array of senior national security and policy officials across the U.S. government, in addition to politicians. The access allowed them to scoop up call logs, unencrypted texts and some audio from targets, in what investigators believe may have significant national-security ramifications.

Additionally, the hackers were able to access information from systems maintained by the carriers to comply with U.S. surveillance requests, raising further counterintelligence concerns. Investigators are still endeavoring to fully understand and have said the attack was carried out by the Salt Typhoon group. At Lumen, which doesn’t provide wireless service, the attackers didn’t steal any customer data or access its wiretap capabilities, according to people familiar with the matter.

Further investigation has revealed that the hackers sought access to data managed under U.S. law enforcement programs, including those governed by the Foreign Intelligence Surveillance Act (FISA).  This act authorizes American intelligence agencies to monitor suspected foreign agents’ communications. By targeting these programs, Chinese hackers may have aimed to infiltrate sensitive government communications channels, gaining insights into U.S. surveillance efforts.

Some foreign telecommunications firms were also compromised in the hacks, including in countries that maintain close intelligence-sharing partnerships with the U.S., people familiar with the matter said.  Earlier this week, the Biden administration acknowledged in a public statement some details about the nature of the “broad and significant” hack that were previously reported by the WSJ.

Chinese government-linked hackers had compromised networks at multiple telecommunications companies “to enable the theft of customer call records data, the compromise of private communications of a limited number of individuals who are primarily involved in government or political activity, and the copying of certain information that was subject to U.S. law enforcement requests pursuant to court orders,” the statement from the FBI and Cybersecurity and Infrastructure Security Agency (CISA) said.  “We expect our understanding of these compromises to grow as the investigation continues,” they added.

References:

https://www.wsj.com/politics/national-security/t-mobile-hacked-in-massive-chinese-breach-of-telecom-networks-4b2d7f92

https://www.wsj.com/tech/cybersecurity/u-s-wiretap-systems-targeted-in-china-linked-hack-327fc63b

https://www.newsweek.com/fbi-chinese-cyber-espionage-multiple-telecom-networks-1985617

China backed Volt Typhoon has “pre-positioned” malware to disrupt U.S. critical infrastructure networks “on a scale greater than ever before”

FBI and MI5 Chiefs Issue Joint Warning: Chinese Cyber Espionage on Tech & Telecom Firms

Cybersecurity threats in telecoms require protection of network infrastructure and availability

FT: A global satellite blackout is a real threat; how to counter a cyber-attack?

Demythifying Cyber security: IEEE ComSocSCV April 19th Meeting Summary

StrandConsult Analysis: European Commission second 5G Cybersecurity Toolbox report

Cisco to lay off more than 4,000 as it shifts focus to AI and Cybersecurity

 

Pente Networks, MosoLabs and Alliance Corp collaborate for Private Cellular Network in a Box

Pente Networks, a private wireless network core, orchestration and management software company; MosoLabs, a global company building solutions for LTE and 5G private and neutral host networks; are collaborating with Alliance Corporation to deliver a complete Private Cellular Network in a Box. It’s a pre-built, plug and play kit with multiple options available (see below) to cater to different use cases, such as surveillance, fixed wireless access and point of sale.

By including already provisioned devices, customers can use their own highly secure and robust private cellular network almost immediately. The solution speeds up and simplifies private network deployment for integrators, IT departments and small businesses; even those with little to no private wireless experience.

Alliance’s Network in a Box is a pre-configured bundle designed with everything needed to deploy and then operate an end-to-end private wireless 4G LTE or 5G wireless network. Alliance’s professional services team uses the Pente HyperCore platform to pre-integrate with MosoLabs’ Canopy indoor and outdoor radios and provision selected end-user devices, SIMs and eSIMs. Alliance then assembles an enclosure that houses all the network equipment, including the HyperCore EdgeMini, a full Evolved Packet Core (EPC) and 5G Core in one. The pre-built kits and radios are ready for simple installation at the desired location. Once connected, the network is immediately up and running, with the devices instantly recognized by the network. No additional programming or configuration is needed.

Three ready made bundles are:

  • Network in a Box with Fixed Wireless Access WiFi
  • Network in a Box with Video Surveillance
  • Network in a Box with Point of Sale

Each bundle includes the Pente HyperCore EdgeMini, Moso Canopy Small Cell Radios, and select devices all preconfigured to work together instantly.

“Our goal is to make private networks accessible to everyone,” says Dan Ortega, Senior Sales Engineer, Alliance Corporation. “With Pente and MosoLabs, we can pre-configure an out-of-the-box functional kit so that all our customers, regardless of their size or use case, can take advantage of everything private networks have to offer.”

Alliance support is available for any questions, and they offer additional services needed such as RF planning, SAS registration, and network monitoring and management. Alliance can also support customers who want to scale their network to support additional services, applications, and IoT devices.

About Alliance:

Alliance Corporation distributes equipment for wireless network infrastructure, in-building signal enhancement solutions, cellular broadband systems, next-generation 5G networks, fixed wireless and private enterprise networks, as well as cellular solutions that connect the Internet of Things. Alliance provides pre and post-sale technical support, engineering, radio configuration, and training services. Alliance serves telecommunication carriers, fixed wireless broadband service providers, OEMs, systems integrators, resellers, and contractors in education, enterprise, federal government, military, healthcare, industrial, municipal government, oil and gas, mining, public safety, security, utilities, and transportation industries.

Alliance US Holding, LP owns Alliance Corporation, GetWireless, TESSCO Technologies, and DiscountCell and is owned by entities affiliated with Lee Equity Partners and Twin Point Capital

About Pente:
Pente provides the world’s most advanced LTE & 5G SA mobile core and network management solution supporting operators and enterprises with both fully virtualized mobile core networks and private cellular network solutions. With over 1000 APIs, Pente’s mobile core and network orchestration platform, enables the fastest implementation and integration available on the market today. Pente currently manages millions of IT-grade SIMs for customers around the globe on networks serving corporate enterprise, industrial, healthcare, education, the public sector and military applications. Pente has strategic customers and partnerships with leading MNOs, MVNOs, enterprises, device manufacturers, managed service providers and system integrators.

Visit pentenetworks.com or call 1-888-821-4797.

About MosoLabs:
MosoLabs is focused on building world-class 4G and 5G hardware with a unified network management platform and an innovative application software suite for private wireless and neutral hosts networks. Our mission is to simplify the entire private network experience – from planning to deployment to management – and create products to support new use cases. We develop fully integrated products that simplify time-to-market and deployment complexity for global enterprises and managed service providers. Learn more at www.mosolabs.com.

References:

https://www.prnewswire.com/news-releases/pente-networks-and-mosolabs-announce-private-cellular-network-in-a-box-options-now-available-through-alliance-corporation-302305633.html

https://www.pentenetworks.com/networkinabox

Tata Consultancy Services: Critical role of Gen AI in 5G; 5G private networks and enterprise use cases

SNS Telecom & IT: $6 Billion Private LTE/5G Market Shines Through Wireless Industry’s Gloom

SNS Telecom & IT: Private 5G Network market annual spending will be $3.5 Billion by 2027

HPE Aruba Launches “Cloud Native” Private 5G Network with 4G/5G Small Cell Radios

Nokia and Kyndryl extend partnership to deliver 4G/5G private networks and MEC to manufacturing companies

BT and Ericsson in partnership to provide commercial 5G private networks in the UK

Ericsson and e& (UAE) sign MoU for 6G collaboration vs ITU-R IMT-2030 framework

Ericsson and United Arab Emirates (UAE) network operator e&  have signed of a Memorandum of Understanding (MoU),  for the collaborative exploration of 6G technology, its use cases and future network evolution.  It will also include a series of technical discussions and engagements aimed at jointly exploring key 6G technology concepts.

The purpose of this MoU is unclear, as the definition work for 6G RANs will be done in ITU-R WP5D with the specs likely to come from 3GPP.  So any 6G MoU would have to be based on the ITU-R IMT-2030 framework (see figures below and References).

Khalid Murshed, Chief Technology and Information Officer, e& UAE, says: “e& UAE has pioneered new technologies since 1976 powering people and societies. This collaboration is a testament to our dedication for driving the digital future and pushing the boundaries of a more connected and  technologically advanced future. We are thrilled to partner with Ericsson on exploring 6G and its future network evolution.”

Ekow Nelson, Vice President and Head of Global Customer Unit for e& at Ericsson Middle East and Africa, says: “We have barely scratched the surface with 5G which will overtake 4G and become the dominant mobile technology after 2027 and, with 5G Standalone and 5G Advanced, realize its transformative potential over the next several years. At the same time, we have started the proactive approach to 6G research with our partners to shape the next generation of mobile networks. Collaborating closely with e& UAE, we aim to leverage our shared expertise to drive progress in the development of 6G for the United Arab Emirates, and the wider region.”

Photo credit: Ericsson

………………………………………………………………………………………………………..

From the ITU-R IMT-2030 framework:

 

References:

https://www.ericsson.com/en/press-releases/5/2024/ericsson-and-eand-uae-sign-mou-for-6g-collaboration

ITU-R: IMT-2030 (6G) Backgrounder and Envisioned Capabilities

ITU-R WP5D invites IMT-2030 RIT/SRIT contributions

Highlights of 3GPP Stage 1 Workshop on IMT 2030 (6G) Use Cases

NGMN issues ITU-R framework for IMT-2030 vs ITU-R WP5D Timeline for RIT/SRIT Standardization

 

FT: New benchmarks for Gen AI models; Neocloud groups leverage Nvidia chips to borrow >$11B

The Financial Times reports that technology companies are rush­ing to redesign how they test and eval­u­ate their Gen AI mod­els, as cur­rent AI bench­marks appear to be inadequate.  AI benchmarks are used to assess how well an AI model can generate content that is coherent, relevant, and creative. This can include generating text, images, music, or any other form of content.

OpenAI, Microsoft, Meta and Anthropic have all recently announced plans to build AI agents that can execute tasks for humans autonom­ously on their behalf. To do this effect­ively, the AI sys­tems must be able to per­form increas­ingly com­plex actions, using reas­on­ing and plan­ning.

Cur­rent pub­lic AI bench­marks — Hel­laswag and MMLU — use mul­tiple-choice ques­tions to assess com­mon sense and know­ledge across vari­ous top­ics. However, research­ers argue this method is now becom­ing redund­ant and mod­els need more com­plex prob­lems.

“We are get­ting to the era where a lot of the human-writ­ten tests are no longer suf­fi­cient as a good baro­meter for how cap­able the mod­els are,” said Mark Chen, senior vice-pres­id­ent of research at OpenAI. “That cre­ates a new chal­lenge for us as a research world.”

The SWE Veri­fied benchmark was updated in August to bet­ter eval­u­ate autonom­ous sys­tems based on feed­back from com­pan­ies, includ­ing OpenAI. It uses real-world soft­ware prob­lems sourced from the developer plat­form Git­Hub and involves sup­ply­ing the AI agent with a code repos­it­ory and an engin­eer­ing issue, ask­ing them to fix it. The tasks require reas­on­ing to com­plete.

“It is a lot more chal­len­ging [with agen­tic sys­tems] because you need to con­nect those sys­tems to lots of extra tools,” said Jared Kaplan, chief sci­ence officer at Anthropic.

“You have to basic­ally cre­ate a whole sand­box envir­on­ment for them to play in. It is not as simple as just provid­ing a prompt, see­ing what the com­ple­tion is and then eval­u­at­ing that.”

Another import­ant factor when con­duct­ing more advanced tests is to make sure the bench­mark ques­tions are kept out of the pub­lic domain, in order to ensure the mod­els do not effect­ively “cheat” by gen­er­at­ing the answers from train­ing data, rather than solv­ing the prob­lem.

The need for new bench­marks has also led to efforts by external organ­iza­tions. In Septem­ber, the start-up Scale AI announced a project called “Human­ity’s Last Exam”, which crowd­sourced com­plex ques­tions from experts across dif­fer­ent dis­cip­lines that required abstract reas­on­ing to com­plete.

Meanwhile, the Financial Times recently reported that Wall Street’s largest financial institutions had loaned more than $11bn to “neocloud” groups, backed by their possession of Nvidia’s AI GPU chips. These companies include names such as CoreWeave, Crusoe and Lambda, and provide cloud computing services to tech businesses building AI products. They have acquired tens of thousands of Nvidia’s graphics processing units (GPUs) through partnerships with the chipmaker. With capital expenditure on data centres surging, in the rush to develop AI models, the Nvidia’s AI GPU chips have become a precious commodity.

Nvidia’s chips have become a precious commodity in the ongoing race to develop AI models © Marlena Sloss/Bloomberg

…………………………………………………………………………………………………………………………………

The $3tn tech group’s allocation of chips to neocloud groups has given confidence to Wall Street lenders to lend billions of dollars to the companies that are then used to buy more Nvidia chips. Nvidia is itself an investor in neocloud companies that in turn are among its largest customers. Critics have questioned the ongoing value of the collateralised chips as new advanced versions come to market — or if the current high spending on AI begins to retract. “The lenders all coming in push the story that you can borrow against these chips and add to the frenzy that you need to get in now,” said Nate Koppikar, a short seller at hedge fund Orso Partners. “But chips are a depreciating, not appreciating, asset.”

References:

https://www.ft.com/content/866ad6e9-f8fe-451f-9b00-cb9f638c7c59

https://www.ft.com/content/fb996508-c4df-4fc8-b3c0-2a638bb96c19

https://www.ft.com/content/41bfacb8-4d1e-4f25-bc60-75bf557f1f21

Tata Consultancy Services: Critical role of Gen AI in 5G; 5G private networks and enterprise use cases

Reuters & Bloomberg: OpenAI to design “inference AI” chip with Broadcom and TSMC

AI adoption to accelerate growth in the $215 billion Data Center market

AI Echo Chamber: “Upstream AI” companies huge spending fuels profit growth for “Downstream AI” firms

AI winner Nvidia faces competition with new super chip delayed

 

Tata Consultancy Services: Critical role of Gen AI in 5G; 5G private networks and enterprise use cases

“In the realm of network management, Generative AI could play a critical role in predicting 5G data flow patterns and optimizing performance, ultimately improving the customer experience. Similarly, in the area of security, Gen AI could be pivotal in identifying and predicting threats before they occur, strengthening overall network security,”  Mayank Gupta, Global Head (Sales and Strategy Network Solutions and Services) at Tata Consultancy Services (TCS), told the Economic Times of India.

 

5G adoption has been happening in stages he said. When asked about the 5G monetization challenges, Gupta opined that one of the main challenges has been fully understanding the business case or use case required by the industry. “It is not the traditional ‘design, deploy and get paid’ model. Instead, you have to engage with the industry, collaborate with various stakeholders and take an ecosystem approach. It is essential to understand the specific business needs, define the problem and then build solutions around that,” Gupta said.

“Two years ago, 5G was moving at a very slow pace, but it has gained significant momentum. Globally, the U.S. and Canada are leading the way, while countries like South Korea and Brazil are accelerating their adoption, largely due to the liberalization of their spectrum. From a deployment standpoint, mobile private networks are among the first areas seeing significant progress, with deployments growing rapidly. “The focus is shifting from just connectivity to a broader transformation of operational technology,” he added.

Over 50% of 5G mobile data is now coming from the enterprise sector, which has become a major focus. Different countries prioritize different use cases based on their needs.

Watch the video interview here:

…………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………………….

References:

https://telecom.economictimes.indiatimes.com/news/enterprise-services/5g-with-genai-to-drive-transformation-across-industries-tcs/115107404

Tata Communications launches global, cloud-based 5G Roaming Lab

5G Made in India: Bharti Airtel and Tata Group partner to implement 5G in India

LightCounting & TÉRAL RESEARCH: India RAN market is buoyant with 5G rolling out at a fast pace

Reliance Jio in talks with Tesla to deploy private 5G network for the latter’s manufacturing plant in India

Communications Minister: India to be major telecom technology exporter in 3 years with its 4G/5G technology stack

India to set up 100 labs for developing 5G apps, business models and use-cases

Adani Group to launch private 5G network services in India this year

Telco spending on RAN infrastructure continues to decline as does mobile traffic growth

Telco spending on radio access network (RAN) infrastructure, which is the largest share of capex, has dropped sharply in the last couple of years. It fell 11% in 2023, to about $40 billion, according to Informa owned market research firm Omdia (see References below for Dell’Oro’s numbers).

For 2024, Omdia predicts another decline of between 7% and 9%.  Instead of buying new equipment, telcos have used existing supplies in their where houses.

Traffic growth rates also appear to have slowed. The graphic used in Ericsson’s latest mobility report pictures this very clearly, showing a church steeple of a traffic spike in 2019 and 2020 before a shallower downward-sloping gradient to the first half of 2024.  That’s shown in this chart:

Source:  Ericsson

Ericsson’s latest numbers, available through its mobility visualizer tool, shows the monthly volume of global mobile data rose by just 4.34% in the second quarter of 2024, compared with the first quarter. This is much lower than the rate of 10.76% Ericsson observed in the corresponding quarter four years earlier. The actual increase in 2024 was 6.27 exabytes. In 2020, it was 4.86 exabytes.

RAN product revenues have been falling at their sharpest rate in many years despite the 6.27-exabyte increase in monthly data volumes that happened in the second quarter. There are no signs that current 4G and 5G networks are about to keel over beneath an avalanche of data.  It remains to be seen whether mobile networks are sufficiently robust to cope with many more exabytes of mobile data traffic or if telcos care about any service problems caused by congestion due to increased traffic.  

There is also no obvious correlation between traffic growth and expenditure, according to Coleago Consulting. Spain’s Telefónica supported 17,054 petabytes of data on its global networks in 2015, its annual reports show. By 2023, the amount had rocketed to 146,074. Yet its capital intensity has fallen from more than 17% to just 14% over this period. Energy use, a proxy for operating costs, is also down, dropping from about 6,578 gigawatt hours in 2015 to 6,012 last year. Despite all that data consumption by its customers, Elisa’s capital intensity last year was less than 15%.

In Germany, the average monthly data usage per mobile customer (rather than per capita) amounted to 7.4 Gbytes and this may have risen to around 8 Gbytes in 1H 2024. In 2024, mobile data consumption in Finland was around 10 times higher compared to Germany.  In Finland mobile operators have implemented 5G-SA and sell a user experience in terms of speed (Mbits/s) as opposed to data volume (Gbytes). As of October 2024, Elisa Finland offered a speed of 300 Mbits/s with unlimited data volume for €34.99. By contrast, in Germany, Telekom’s offer for 20 Gbytes is priced at €39.95 per month. For unlimited data usage Telekom charges €84.95, which is 2.5 times more costly than Elisa’s unlimited offer. It is unreasonable to assume that there is no price elasticity of demand. Surely, if prices in Germany were like those in Finland, monthly mobile data usage per customer would be much higher.

Monthly average revenue per user (ARPU) for a postpaid customer of Deutsche Telekom, Germany’s biggest telco, has fallen from €22 (US$23.7) before the launch of 5G to about €20 ($21.6) for the most recent quarter.  Clearly, cost realities are especially awkward for Europe’s telcos, which have refused to give up their “fair share” argument that big content companies should pay for network usage because of all the traffic they supposedly generate. Critics disagree, saying that the telco’s own customers are the traffic generators, and they have already paid for it, even if pricing schemes do not help telcos to grow their sales.

………………………………………………………………………………………………………………………………………………………………………………………………………………………………..

References:

https://www.lightreading.com/5g/data-traffic-growth-or-decline-there-s-no-upside-for-telecom

https://www.ericsson.com/en/reports-and-papers/mobility-report/mobility-visualizer?f=1&ft=2&r=2,3,4,5,6,7,8,9&t=1,2,3,4,5,6,7&s=4&u=1&y=2023,2029&c=3

https://www.coleago.com/insights/the-end-of-telecoms-history-not-really/

Analysys Mason & Light Reading: cellular data traffic growth rates are decreasing

Dell’Oro: Global RAN Market to Drop 21% between 2021 and 2029

Dell’Oro: RAN market still declining with Huawei, Ericsson, Nokia, ZTE and Samsung top vendors

Dell’Oro & Omdia: Global RAN market declined in 2023 and again in 2024

Dell’Oro: RAN market declines at very fast pace while Mobile Core Network returns to growth in Q2-2023

China’s mobile data consumption slumps; Apple’s market share shrinks-no longer among top 5 vendors

Analysts: Telco CAPEX crash looks to continue: mobile core network, RAN, and optical all expected to decline

Ookla: T-Mobile and Verizon lead in U.S. 5G FWA

 

 

T‑Mobile achieves record 5G Uplink speed with 5G NR Dual Connectivity

T-Mobile US claims it broke a world record with its 5G standalone (SA) network via a new feature called New Radio Dual Connectivity (5G DC) [1.]. With 5G DC.  The so called “Un-carrier” was able to massively increase uplink throughput and capacity, reaching peak speeds of 2.2 Gbps — that’s the fastest recorded anywhere in the world — and demonstrates the technology’s potential to create serious efficiencies in how data is transmitted from devices to the network.

Note 1. New Radio Dual Connectivity (NR-DC) is a dual connectivity configuration that uses the 5G standalone core (specified by 3GPP but not standardized by ITU-R or ITU-T). In this configuration, both the primary and secondary RAN nodes are 5G gNBs.  NR-DC was was specified in 3GPP Release 15 along with simultaneous receive (Rx) / transmit (Tx) band combinations for NR CA/DC.

…………………………………………………………………………………………………………………………………..

To put T-Mo’s  2.2 Gbps uplink speed into context, the latest report from connectivity data specialist Ookla puts the median mobile upload speed in the U.S. at 8.41 Mbps, although that’s across networks. T-Mobile is ahead of major rivals AT&T and Verizon with a median upload speed of 12.19 Mbps.

In June Ookla stated that while U.S. network operators have invested heavily in improving 5G download speeds, “5G upload and latency performance need more attention.” Its data at the time showed Verizon and T-Mobile had comparable 5G upload at just above 15 Mbps, while AT&T lagged somewhat at closer to the 10 Mbps mark.

5G DC enables the Un-carrier to aggregate 2.5 GHz and mmWave spectrum, allowing for an insane boost to uplink throughput and capacity. In this test, T-Mobile was able to allocate 60% of the mmWave radio resources for uplink where previous use cases typically allowed up to 20%. Completed on T-Mobile’s 5G SA production network in SoFi Stadium in Southern California with equipment and 5G DC solution from Ericsson and a mobile test smartphone powered by a flagship Snapdragon® X80 5G Modem-RF System from Qualcomm Technologies, Inc., this test changes the game for providers looking to offer customers and businesses the best experience possible at crowded events.

“With 5G DC, T-Mobile is pushing the boundaries of what’s possible to create better experiences in the places that matter most to our customers,” said Ulf Ewaldsson, President of Technology at T-Mobile. “This accomplishment is a testament to the network we’ve built over the last five years and our ability to deliver unparalleled capabilities that extend beyond the devices in our pockets.”

For those in the know, download speeds typically reign as the top network performance metric, but with recent strides in uplink capabilities and increasingly demanding tasks, upload speed is becoming more important than ever, especially for live events, mobile gaming and extended reality applications.

Because of this, SoFi Stadium served as the perfect test site for 5G DC. Every year, millions of people flock to the stadium for the latest football game or to catch their favorite artists in concert. Naturally, all these people want to post, livestream and share their experiences in real-time, which can sometimes be a challenge at crowded events with limited capacity. Not to mention broadcast crews who need to upload high-definition content to production teams in real-time for those watching at home. With 5G DC and T-Mobile, all of this gets done faster than ever, alleviating posting FOMO and production crew headaches.

Mårten Lerner, Head of Product Area Networks at Ericsson, said: “High uplink speeds are essential for delivering immersive experiences and reliable 5G connectivity. This mirrors one of our key objectives with the recent launch of Ericsson 5G Advanced, which is to elevate user experience by enhancing network performance for more interactive applications. This 5G uplink speed milestone, achieved with T-Mobile and Qualcomm, underscores our commitment to taking user experience to unprecedented levels.”

“We are incredibly proud to achieve yet another world record with T-Mobile. This groundbreaking achievement shows what could be possible with 5G DC and how it can bring new, unparalleled experiences to consumers, especially at large events like football games and concerts,” said Sunil Patil, Vice President, Product Management, Qualcomm Technologies, Inc. “We will continue our close collaboration with global innovators like T-Mobile and Ericsson to push the boundaries and unlock the full potential of 5G.”

5G network covers more than 330 million people across two million square miles. More than 300 million people nationwide are covered by T-Mobile’s super-fast Ultra Capacity 5G with over 2x more square miles of coverage than similar mid-band 5G offerings from the Un-carrier’s closest competitors.

For more information on T-Mobile’s network, visit T-Mobile.com/coverage.

References:

https://www.t-mobile.com/news/network/t-mobile-shatters-for-5g-uplink-speed

https://www.telecoms.com/5g-6g/t-mobile-us-uses-5g-dc-to-claim-uplink-speed-record

Telstra achieves 340 Mbps uplink over 5G SA; Deploys dynamic network slicing from Ericsson

Finland’s Elisa, Ericsson and Qualcomm test uplink carrier aggregation on 5G SA network

Ericsson and MediaTek set new 5G uplink speed record using Uplink Carrier Aggregation

Samsung-Mediatek 5G uplink trial with 3 transmit antennas

Dish Wireless with Qualcomm Technologies and Samsung test simultaneous 5G 2x uplink and 4x downlink carrier aggregation

BT, Nokia and Qualcomm demonstrate 2CC CA on uplink of a 5G SA network

 

 

Page 2 of 323
1 2 3 4 323