Author: Alan Weissberger
Ericsson integrates agentic AI into its NetCloud platform for self healing and autonomous 5G private networks
Ericsson is integrating agentic AI into its NetCloud platform to create self-healing and autonomous 5G private (enterprise) networks. This initiative upgrades the existing NetCloud Assistant (ANA), a generative AI tool, into a strategic partner capable of managing complex workflows and orchestrating multiple AI agents. The agentic AI agent aims to simplify private 5G adoption by reducing deployment complexity and the need for specialized administration. This new agentic architecture allows the new Ericsson system to interpret high-level instructions and autonomously assign tasks to a team of specialized AI agents.
Key AI features include:
- Agentic organizational hierarchy: ANA will be supported by multiple orchestrator and functional AI agents capable of planning and executing (with administrator direction). Orchestrator agents will be deployed in phases, starting with a troubleshooting agent planned in Q4 2025, followed by configuration, deployment, and policy agents planned in 2026. These orchestrators will connect with task, process, knowledge, and decision agents within an integrated agentic framework.
- Automated troubleshooting: ANA’s troubleshooting orchestrator will include automated workflows that address the top issues identified by Ericsson support teams, partners, and customers, such as offline devices and poor signal quality. Planned to launch in Q4 2025, this feature is expected to reduce downtime and customer support cases by over 20 percent.
- Multi-modal content generation: ANA can now generate dynamic graphs to visually represent trends and complex query results involving multiple data points.
- Explainable AI: ANA displays real-time process feedback, revealing steps taken by AI agents in order to enhance transparency and trust.
- Expanded AIOps insights: NetCloud AIOps will be expanded to provide isolation and correlation of fault, performance, configuration, and accounting anomalies for Wireless WAN and NetCloud SASE. For Ericsson Private 5G, NetCloud is expected to provide service health analytics including KPI monitoring and user equipment connectivity diagnostics. Planned availability Q4 2025.
Manish Tiwari, Head of Enterprise 5G, Ericsson Enterprise Wireless Solutions, adds: “With the integration of Ericsson Private 5G into the NetCloud platform, we’re taking a major step forward in making enterprise connectivity smarter, simpler, and adaptive. By building on powerful AI foundations, seamless lifecycle management, and the ability to scale securely across sites, we are providing flexibility to further accelerate digital transformation across industries. This is about more than connectivity: it is about giving enterprises the business-critical foundation they need to run IT and OT systems with confidence and unlock the next wave of innovation for their businesses.”
Pankaj Malhotra, Head of WWAN & Security, Ericsson Enterprise Wireless Solutions, says: “By introducing agentic AI into NetCloud, we’re enabling enterprises to simplify deployment and operations while also improving reliability, performance, and user experience. More importantly, it lays the foundation for our vision of fully autonomous, self-optimizing 5G enterprise networks, that can power the next generation of enterprise innovation.”
Agentic AI and the Future of Communications for Autonomous Vehicle (V2X)
Ericsson completes Aduna joint venture with 12 telcos to drive network API adoption
Ericsson reports ~flat 2Q-2025 results; sees potential for 5G SA and AI to drive growth
Ericsson revamps its OSS/BSS with AI using Amazon Bedrock as a foundation
Ericsson’s sales rose for the first time in 8 quarters; mobile networks need an AI boost
Elon Musk: Starlink could become a global mobile carrier; 2 year timeframe for new smartphones
Yesterday, during a segment of the All-in Podcast dedicated to the SpaceX-EchoStar spectrum sales agreement [1.], Space X/Starlink boss Elon Musk was asked if this sets the industry down a path where Starlink’s end goal is to emerge as a global carrier that, effectively, would limit the role of regional carriers. “That would be one of the options,” Musk responded. Musk downplayed any threat against AT&T, Verizon and T-Mobile. The podcast section dedicated to the EchoStar agreement starts around the 16:50 mark. You can start watching at that point via this YouTube link.
Note 1. SpaceX’s $17 billion agreement with EchoStar includes $8.5 billion in stock, plus $2 billion of cash interest payments payable on EchoStar debt. Separately, AT&T’s is paying $23 billion – all in cash – for its acquisition of EchoStar’s spectrum.
Regarding the EchoStar spectrum deal, Musk said, “This is kind of a long term thing. It will allow SpaceX to deliver high bandwidth connectivity directly from the satellites to the phones.”
Musk said that deal would not seriously challenge the big three U.S. mobile carriers. He said:
“To be clear, we’re not going to put the other carriers out of business. They’re still going to be around because they own a lot of spectrum. But, yes, you should be able to have a Starlink, like you have an AT&T or T-Mobile or Verizon, or whatever. You can have an account with Starlink that works with your Starlink [satellite] antenna at home with … Wi-Fi, as well as on your phone. We’d be a comprehensive solution for high bandwidth at home and high bandwidth for direct-to-cell.”
“Could you buy Verizon?” Musk was asked. “Not out of the question. I suppose that may happen,” Musk said with a chuckle.
That idea at least “highlights the possibility that SpaceX could pursue additional spectrum,” LightShed Partners analysts Walter Piecyk and Joe Galone explained in this blog post. “We highly doubt SpaceX has any interest in the people or infrastructure of a telco, there are plenty of compelling spectrum assets in and outside of those carriers to consider.”
Getting smartphones equipped with chips to support those new frequency bands will take some time. Musk estimated that’s “probably a two-year timeframe.” LightShed Partners analysts agreed, “On devices, Elon’s two-year timeline for a Starlink phone isn’t surprising given spectrum banding, chip development, and satellite integration. He’s mused before that if phone manufacturers continued to hinder his technology that he “would make a phone as a forcing function to compete with them.”
Some analysts view MVNO agreements as Starlink’s best route to becoming a full scale mobile carrier of satellite and terrestrial wireless services.
“The most plausible business model is that Starlink partners with MNOs for them to resell the service or embed the service as part of their plans,” Lluc Palerm Serra, research director at Analysys Mason, told PCMag.
LightShed Partners agreed. Musk’s point that SpaceX isn’t out to displace the incumbent carriers “reinforced our view that securing an MVNO deal will be essential if SpaceX wants to deliver a Starlink phone directly to consumers,” LightShed’s Walter Piecyk and Joe Galone explained in this blog post.
“In parallel, we’re working on the satellites and working with the handset makers to add these frequencies to the phones,” Musk said. “And the phones will then handshake well to achieve high-bandwidth connectivity. The net effect is that you should be able to watch videos anywhere on your phone.”
AT&T CEO John Stankey addressed Starlink’s “mobile-first” possibility earlier this week at an investor conference. Starlink’s current access to spectrum, including what is coming way of EchoStar, isn’t enough to create a “robust terrestrial replacement,” he said. But he acknowledged that, with the right type of commitments, perhaps it could happen someday.
“EchoStar still owns the highly lucrative 700 and AWS-3 spectrum, in which we note that all three wireless carriers have a robust ecosystem,” TD Cowen analyst Gregory Williams wrote in a note earlier this week. “Whether EchoStar sells more [spectrum] in short order remains to be seen,” TD Cowen’s Williams wrote Monday, explaining that, with the FCC dispute resolved, it may hold onto its portfolio longer. “EchoStar is not a forced seller, now has an excellent balance sheet and liquidity, and may desire to hold onto the spectrum as long as possible for higher sale valuations at a later date,” he added.
References:
https://www.lightreading.com/5g/turning-starlink-into-a-global-carrier-one-of-the-options-musk-says
Elon Talks Starlink Phone. Disruption Looms for Telcos and Apple
Huge significance of EchoStar’s AWS-4 spectrum sale to SpaceX
U.S. BEAD overhaul to benefit Starlink/SpaceX at the expense of fiber broadband providers
Telstra selects SpaceX’s Starlink to bring Satellite-to-Mobile text messaging to its customers in Australia
SpaceX launches first set of Starlink satellites with direct-to-cell capabilities
SpaceX has majority of all satellites in orbit; Starlink achieves cash-flow breakeven
Highlights of Nokia’s Smart Factory in Oulu, Finland for 5G and 6G innovation
Nokia has opened a Smart Factory in Oulu, Finland, for 5G/6G design, manufacturing, and testing, integrating AI technologies and Industry 4.0 applications. It brings ~3,000 staff under one roof and is positioned as Europe’s flagship site for radio access (RAN) innovation.
The Oulu campus will initially focus on 5G, including: Standardization, System-on Chips as well as 5G radio hardware and software and patents. Oulu Factory, part of the new campus, will target New Production Introduction for Nokia’s 5G radio and baseband products. The new campus strengthens Oulu’s ecosystem as a global testbed for resilient and secure networks for both civilian and defense applications.
At Oulu “Home of Radio” campus, Nokia’s research and innovation underpins high quality, tested world class products readymade for customers across markets. Nokia’s experts will continue to foster innovation, from Massive MIMO radios like Osprey and Habrok to next-generation 6G solutions, creating secure, high-performance, future-proof connectivity.
Sustainability is integral to the facility. Renewable energy is used throughout the site, with additional energy used to heat 20,000 households in Oulu. The on-site energy station is one of the world’s largest CO2-based district heating and cooling plants.
Active 6G proof-of-concept trials will be tested using ~7 GHz and challenging propagation scenarios.
“Our teams in Oulu are shaping the future of 5G and 6G developing our most advanced radio networks. Oulu has a unique ecosystem that integrates Nokia’s R&D and smart manufacturing with an ecosystem of partners – including universities, start-ups and NATO’s DIANA test center. Oulu embodies our culture of innovation and the new campus will be essential to advancing connectivity necessary to power the AI supercycle,” said Justin Hotard, President and CEO of Nokia
Nokia Oulu Facts:
- Around 3,000 employees and 40 nationalities working on the campus.
- Oulu campus covers the entire product lifecycle of a product, from R&D to manufacturing and testing of the products.
- Footprint of the building is overall 55,000 square metres, including manufacturing, R&D and office space.
- Green campus with all energy purchased green and all surplus energy generated fed back into the district heating system and used to heat 20,000 local households.
- The campus boasts 100% waste utilization rate and 99% avoidance in CO2 emissions.
- Construction started in the second half of 2022, with the first employees moving into the facility in the first half of this year.
- YIT constructed the site and Arkkitehtitoimisto ALA were the architects.
References:
https://www.sdxcentral.com/analysis/behind-the-scenes-at-nokias-new-home-of-radio/
Will the wave of AI generated user-to/from-network traffic increase spectacularly as Cisco and Nokia predict?
Nokia’s Bell Labs to use adapted 4G and 5G access technologies for Indian space missions
Indosat Ooredoo Hutchison and Nokia use AI to reduce energy demand and emissions
Verizon partners with Nokia to deploy large private 5G network in the UK
Nokia selects Intel’s Justin Hotard as new CEO to increase growth in IP networking and data center connections
Nokia sees new types of 6G connected devices facilitated by a “3 layer technology stack”
Nokia and Eolo deploy 5G SA mmWave “Cloud RAN” network
Huge significance of EchoStar’s AWS-4 spectrum sale to SpaceX
EchoStar has entered into a definitive agreement to sell its entire portfolio of prized AWS-4 [1.] and H-block spectrum licenses to SpaceX in a deal valued at approximately $19 billion. The spectrum purchase allows SpaceX to start building and deploying upgraded, laser-connected satellites that the company said will expand the cell network’s capacity by “more than 100 times.”
This deal marks EchoStar/Dish Network’s exit as a mobile network provider (goodbye multi-vendor 5G OpenRAN) which once again makes the U.S. wireless market a three-player (AT&T, Verizon, T-Mobile) affair. Despite that operational failure, the deal helps EchoStar address regulatory pressure and strengthen its financial position, especially after AT&T agreed to buy spectrum licenses from EchoStar for $23 billion. 
The companies also agreed to a deal that will enable EchoStar’s Boost Mobile subscribers to access Starlink direct-to-cell (D2C) service to extend satellite service to areas without mobile network service.
……………………………………………………………………………………………………………………………………………………………
Note 1. The AWS-4 spectrum band (2000-2020 MHz and 2180-2200 MHz) is widely considered the “golden band” for D2C services. Unlike repurposed terrestrial spectrum, the AWS-4 band was originally allocated for Mobile Satellite Service (MSS).
…………………………………………………………………………………………………………………………………………………………..
The AWS-4 spectrum acquisition transforms SpaceX from a D2C partner into an owner that controls its dedicated MSS spectrum. The deal with EchoStar will allow SpaceX to operate Starlink direct-to-cell (D2C) services on frequencies it owns, rather than relying solely on those leased from mobile carriers like T-Mobile and other mobile operators it’s working with (see References below).
Roger Entner wrote that SpaceX is now a “kingmaker.” He emailed this comment:
“With 50 MHz of dedicated spectrum, the raw bandwidth that Starlink can deliver increases by 1.5 GBbit/s. This is a substantial increase in speed to customers. The math is 30 bit/s/hz which is LTE spectral efficiency x 50 MHz = 1.5 Gbit/s. “This agreement makes Starlink an even more serious play in the D2C market as it will have first hand experience with how to utilize terrestrial spectrum. It is one thing to have this experience through a partner, this a completely different game when you own it.”
The combination of T-Mobile’s terrestrial network and Starlink’s enhanced D2C capabilities allows T-Mobile to market a service with virtually seamless connectivity that eliminates outdoor dead zones using Starlink’s spectrum.
“For the past decade, we’ve acquired spectrum and facilitated worldwide 5G spectrum standards and devices, all with the foresight that direct-to-cell connectivity via satellite would change the way the world communicates,” said Hamid Akhavan, president & CEO, EchoStar. “This transaction with SpaceX continues our legacy of putting the customer first as it allows for the combination of AWS-4 and H-block spectrum from EchoStar with the rocket launch and satellite capabilities from SpaceX to realize the direct-to-cell vision in a more innovative, economical and faster way for consumers worldwide.”
“We’re so pleased to be doing this transaction with EchoStar as it will advance our mission to end mobile dead zones around the world,” said Gwynne Shotwell, president & COO, SpaceX. “SpaceX’s first generation Starlink satellites with Direct to Cell capabilities have already connected millions of people when they needed it most – during natural disasters so they could contact emergency responders and loved ones – or when they would have previously been off the grid. In this next chapter, with exclusive spectrum, SpaceX will develop next generation Starlink Direct to Cell satellites, which will have a step change in performance and enable us to enhance coverage for customers wherever they are in the world.”
EchoStar anticipates this transaction with SpaceX along with the previously announced spectrum sale will resolve the Federal Communications Commission’s (FCC) inquiries. Closing of the proposed transaction will occur after all required regulatory approvals are received and other closing conditions are satisfied.
The EchoStar-Space X transaction is structured with a balanced mix of cash and equity plus interest payments:
-
Cash and Stock Components: SpaceX will provide up to $8.5 billion in cash and an equivalent amount in its own stock, with the valuation fixed at the time the agreement was signed. This 50/50 structure provides EchoStar with immediate liquidity to address its creditors while allowing SpaceX to preserve capital for its immense expenditures on Starship and Starlink development. A pure stock deal would have been untenable for EchoStar, which is saddled with over $26.4 billion in debt, while a pure cash deal would have strained SpaceX.
-
Debt Servicing: In a critical provision underscoring EchoStar’s dire financial state, SpaceX has also agreed to fund approximately $2 billion of EchoStar’s cash interest payments through November 2027.
-
Commercial Alliance: The deal establishes a long-term commercial partnership wherein EchoStar’s Boost Mobile subscribers will gain access to SpaceX’s next-generation Starlink D2C service. This provides a desperately needed lifeline for the struggling Boost brand. More strategically, this alliance serves as a masterful piece of regulatory maneuvering. It allows regulators to plausibly argue that they have preserved a “fourth wireless competitor,” providing the political cover necessary to approve a deal that permanently cements a three-player terrestrial market.
The move comes amid rapidly increasing U.S. mobile data usage. In 2024, Americans used a record 132 trillion megabytes of mobile data, up 35% over the prior all-time record, industry group CTIA said Monday.
About EchoStar Corporation:
EchoStar Corporation (Nasdaq: SATS) is a premier provider of technology, networking services, television entertainment and connectivity, offering consumer, enterprise, operator and government solutions worldwide under its EchoStar®, Boost Mobile®, Sling TV, DISH TV, Hughes®, HughesNet®, HughesON™, and JUPITER™ brands. In Europe, EchoStar operates under its EchoStar Mobile Limited subsidiary and in Australia, the company operates as EchoStar Global Australia. For more information, visit www.echostar.com and follow EchoStar on X (Twitter) and LinkedIn.
©2025 EchoStar, Hughes, HughesNet, DISH and Boost Mobile are registered trademarks of one or more affiliate companies of EchoStar Corp.
About SpaceX:
SpaceX designs, manufactures, and launches the world’s most advanced rockets and spacecraft. The company was founded in 2002 to revolutionize space technology, with the ultimate goal of making life multiplanetary. As the world’s leading provider of launch services, SpaceX is leveraging its deep experience with both spacecraft and on-orbit operations to deploy the world’s most advanced internet and Direct to Cell networks. Engineered to end mobile dead zones around the world, Starlink’s satellites with Direct to Cell capabilities enable ubiquitous access to texting, calling, and browsing wherever you may be on land, lakes, or coastal waters.
………………………………………………………………………………………………………………………………………………………………………………………………
References:
Mulit-vendor Open RAN stalls as Echostar/Dish shuts down it’s 5G network leaving Mavenir in the lurch
AT&T to buy spectrum Licenses from EchoStar for $23 billion
SpaceX launches first set of Starlink satellites with direct-to-cell capabilities
Starlink’s Direct to Cell service for existing LTE phones “wherever you can see the sky”
Space X “direct-to-cell” service to start in the U.S. this fall, but with what wireless carrier? (T-Mobile)
KDDI unveils AU Starlink direct-to-cell satellite service
Starlink Direct to Cell service (via Entel) is coming to Chile and Peru be end of 2024
Telstra selects SpaceX’s Starlink to bring Satellite-to-Mobile text messaging to its customers in Australia
Telstra partners with Starlink for home phone service and LEO satellite broadband services
One NZ launches commercial Satellite TXT service using Starlink LEO satellites
Ookla: Global performance of Apple’s in-house designed C1 modem in iPhone 16e
Apple designed the C1 modem entirely in-house [1.] as part of its Apple Silicon efforts, a significant move to reduce its reliance on external vendors like Qualcomm. The C1 modem is featured in the iPhone 16e and is manufactured by TSMC, which produces the chip on 4nm and 7nm process nodes. The C1 modem supports all the low and mid-band 5G spectrum but it doesn’t support 5G mmWave spectrum. It also supports Wi-Fi 6 with 2×2 MIMO and Bluetooth 5.3, but lacks Wi-Fi 7 support unlike the rest of the iPhone 16 series of devices.
Note 1. In 2019, Apple purchased Intel’s smartphone modem business for $1 billion with the explicit goal of eventually designing its own modems. The C1 modem is the first major outcome of this strategic shift, replacing Qualcomm’s modems in certain Apple devices, starting with the iPhone 16e. Historically, Apple relied upon Qualcomm to provide most of its iPhone modems so its decision to use the C1 modem in the iPhone 16e is considered a significant move.
Ookla has just analyzed the performance of the iPhone 16e and compared it to the performance of the iPhone 16 on 5G, using Speedtest Intelligence data for Q2 and Q3 2025.
A few important takeaways include:
- The iPhone 16e with the Apple C1 modem performs similarly to the iPhone 16 with the Qualcomm modem in the vast majority of markets we examined.
- The iPhone 16 with Qualcomm modem performs better on more capable mobile networks that have a 5G standalone (SA) footprint supporting higher carrier aggregation combinations and uplink MIMO technology. The iPhone 16e with the C1 modem is not able to achieve the same frontier of performance in these markets due to its technical limitations. Key examples of networks facilitating stronger performance for the iPhone 16 include those in Saudi Arabia, China, India and the U.S.
- In the U.S., T-Mobile users experienced better performance on the iPhone 16, which supports four-carrier aggregation, than iPhone 16e users with the Apple C1 modem, which supports a maximum of three-carrier aggregation. Median download speed for the iPhone 16 on T-Mobile’s network was 317.64 Mbps, compared to 252.80 Mbps on the iPhone 16e.
- The iPhone 16e performs strongly on other key performance metrics. Across the markets analyzed, it tended to record better download speeds among the 10th percentile of users (those with the lowest overall download speeds), and across 10th, median and 90th percentiles for upload speeds. At the lower 10th percentile it’s likely that more users are connected solely to low-band spectrum (sub GHz) which offers better coverage but slower speeds.
Japan is the most popular market for the iPhone 16e, with 11.3% of samples from the 16 lineup, followed chiefly by European markets. Adoption of the iPhone 16e depends on a range of factors, including the level of subsidies within a market and to which devices they are directed, level of price sensitivity among consumers, as well as launch timing, and consumer preferences for different form factors and device features.
The combination of these factors likely explains the relatively higher 16e penetration observed in Japan. Beyond the historic appetite for lower-cost, compact iPhones like the SE (to which the 16e is a spiritual successor) and a subsidy structure that favors entry variants, the recent weakness of the yen has made the Pro and Pro Max models more expensive in local terms, prompting elastic buyers (like students and families) to shift down the line-up.
……………………………………………………………………………………………………………………………………………………………………………….
The iPhone 16e outperforms the iPhone 16 in median upload speed in 15 of the 21 markets we examined. Canada is perhaps the most dramatic example where iPhone 16e median upload speeds of 23.91 Mbps are more than double the iPhone 16’s median upload speed of 11.57 Mbps. However, once again we saw the iPhone 16 perform strongly in median upload speed in countries with advanced 5G networks such as Saudi Arabia and China. Although in the US market the iPhone 16e outperformed the iPhone 16 in upload speeds, when we drilled down further (see the US section of this report), we found that upload performance varied between the different operators.
…………………………………………………………………………………………………………………………………………………………………………….
In the U.S., the iPhone 16 performs better than the iPhone 16e in median download speed for T-Mobile and Verizon customers. This is a slight change from our March 2025 analysis when the iPhone 16e performed better for Verizon customers than the iPhone 16. Because the iPhone 16 supports mmWave spectrum and mmWave is part of Verizon’s 5G Ultra Wideband service, it’s likely that this is a contributing factor in the iPhone 16’s better performance on the Verizon network. Verizon users on the iPhone 16 only clocked a median download speed of 172.12 Mbps, which is significantly lower than iPhone 16 users on T-Mobile’s network that logged a median download speed of 317.64 Mbps.
When comparing the median upload speeds of the iPhone 16 and 16e across US providers there’s a much different story than when comparing median download speeds. On Verizon’s and AT&T’s networks the iPhone 16e outperforms the iPhone 16 in upload speeds. Verizon iPhone 16e users experienced median upload speeds of 11.51 Mbps compared to Verizon iPhone 16 users that logged median upload speeds of 9.67 Mbps. Likewise, AT&T iPhone 16e users experienced median upload speeds of 8.47 Mbps compared to iPhone 16 users with median upload speeds of 7.09 Mbps.
Instead of seeing the iPhone 16 outperform the iPhone 16e at T-Mobile, the two devices are nearly equal in median UL performance with 16e users seeing median upload speeds of 11.79 Mbps compared to iPhone 16 users with 11.70 Mbps. These results are very similar to what we uncovered in our March 2025 report where we saw clear differences in the iPhone 16e and the iPhone 16 performance for AT&T and Verizon users but nearly equal performance for T-Mobile users.
References:
https://www.ookla.com/articles/iphone-c1-modem-performance-q2-q3-2025
Apple in advanced talks to buy Intel’s 5G modem business for $1 billion
China’s mobile data consumption slumps; Apple’s market share shrinks-no longer among top 5 vendors
AST SpaceMobile completes 1st ever LEO satellite voice call using AT&T spectrum and unmodified Samsung and Apple smartphones
OpenAI and Broadcom in $10B deal to make custom AI chips
Overview:
Late last October, IEEE Techblog reported that “OpenAI the maker of ChatGPT, was working with Broadcom to develop a new artificial intelligence (AI) chip focused on running AI models after they’ve been trained.” On Friday, the WSJ and FT (on-line subscriptions required) separately confirmed that OpenAI is working with Broadcom to develop custom AI chips, a move that could help alleviate the shortage of powerful processors needed to quickly train and release new versions of ChatGPT. OpenAI plans to use the new AI chip internally, according to one person close to the project, rather than make them available to external customers.
………………………………………………………………………………………………………………………………………………………………………………………….
Broadcom:
During its earnings call on Thursday, Broadcom’s CEO Hock Tan said that it had signed up an undisclosed fourth major AI developer as a custom AI chip customer, and that this new customer had committed to $10bn in orders. While Broadcom did not disclose the names of the new customer, people familiar with the matter confirmed OpenAI was the new client. Broadcom and OpenAI declined to comment, according to the FT. Tan said the deal had lifted the company’s growth prospects by bringing “immediate and fairly substantial demand,” shipping chips for that customer “pretty strongly” starting next year. “The addition of a fourth customer with immediate and fairly substantial demand really changes our thinking of what 2026 would be starting to look like,” Tan added.
Image credit: © Dado Ruvic/Reuters
HSBC analysts have recently noted that they expect to see a much higher growth rate from Broadcom’s custom chip business compared with Nvidia’s chip business in 2026. Nvidia continues to dominate the AI silicon market, with “hyperscalers” still representing the largest share of its customer base. While Nvidia doesn’t disclose specific customer names, recent filings show that a significant portion of their revenue comes from a small number of unidentified direct customers, which likely are large cloud providers like Microsoft, Amazon, Alphabet (Google), and Meta Platforms.
In August, Broadcom launched its Jericho networking chip, which is designed to help speed up AI computing by connecting data centers as far as 60 miles apart. By August, Broadcom’s market value had surpassed that of oil giant Saudi Aramco, making the chip firm the world’s seventh-largest publicly listed company.
……………………………………………………………………………………………………………………………………….
Open AI:
OpenAI CEO Sam Altman has been saying for months that a shortage of graphics processing units, or GPUs, has been slowing his company’s progress in releasing new versions of its flagship chatbot. In February, Altman wrote on X that ChatGPT-4.5, its then-newest large language model, was the closest the company had come to designing an AI model that behaved like a “thoughtful person,” but there were very high costs that came with developing it. “We will add tens of thousands of GPUs next week and roll it out to the plus tier then. (hundreds of thousands coming soon, and i’m pretty sure y’all will use every one we can rack up.)”
In recent years, OpenAI has relied heavily on so-called “off the shelf” GPUs produced by Nvidia, the biggest player in the chip-design space. But as demand from large AI firms looking to train increasingly sophisticated models has surged, chip makers and data-center operators have struggled to keep up. The company was one of the earliest customers for Nvidia’s AI chips and has since proven to be a voracious consumer of its AI silicon.
“If we’re talking about hyperscalers and gigantic AI factories, it’s very hard to get access to a high number of GPUs,” said Nikolay Filichkin, co-founder of Compute Labs, a startup that buys GPUs and offers investors a share in the rental income they produce. “It requires months of lead time and planning with the manufacturers.”
To solve this problem, OpenAI has been working with Broadcom for over a year to develop a custom chip for use in model training. Broadcom specializes in what it calls XPUs, a type of semiconductor that is designed with a particular application—such as training ChatGPT—in mind.
Last month, Altman said the company was prioritizing compute “in light of the increased demand from [OpenAI’s latest model] GPT-5” and planned to double its compute fleet “over the next 5 months.” OpenAI also recently struck a data-center deal with Oracle that calls for OpenAI to pay more than $30 billion a year to the cloud giant, and signed a smaller contract with Google earlier this year to alleviate computing shortages. It is also embarking on its own data-center construction project, Stargate, though that has gotten off to a slow start.
OpenAI’s move follows the strategy of tech giants such as Google, Amazon and Meta, which have designed their own specialized custom chips to run AI workloads. The industry has seen huge demand for the computing power to train and run AI models.
………………………………………………………………………………………………………………………………………………………………………………..
References:
https://www.ft.com/content/e8cc6d99-d06e-4e9b-a54f-29317fa68d6f
https://www.wsj.com/tech/ai/openai-broadcom-deal-ai-chips-5c7201d2
Reuters & Bloomberg: OpenAI to design “inference AI” chip with Broadcom and TSMC
Open AI raises $8.3B and is valued at $300B; AI speculative mania rivals Dot-com bubble
OpenAI announces new open weight, open source GPT models which Orange will deploy
OpenAI partners with G42 to build giant data center for Stargate UAE project
Generative AI Unicorns Rule the Startup Roost; OpenAI in the Spotlight
Will the wave of AI generated user-to/from-network traffic increase spectacularly as Cisco and Nokia predict?
Network operators are bracing themselves for a wave of AI traffic, partially based on a RtBrick survey, as well as forecasts by Cisco and Nokia, but that hasn’t happened yet. The heavy AI traffic today is East to West (or vice-versa) within cloud resident AI data centers and for AI data center interconnects.
1. Cisco believes that AI Inference agents will soon engage “continuously” with end-users, keeping traffic levels consistently high. has stated that AI will greatly increase network traffic, citing a shift toward new, more demanding traffic patterns driven by “agentic AI” and other applications. This perspective is a core part of Cisco’s business strategy, which is focused on selling the modernized infrastructure needed to handle the coming surge. Cisco identified three stages of AI-driven traffic growth, each with different network demands:
- Today’s generative AI models produce “spikey” traffic which spikes up when a user submits a query, but then returns to a low baseline. Current networks are largely handling this traffic without issues.
- Persistent “agentic” AI traffic: The next phase will involve AI agents that constantly interact with end-users and other agents. Cisco CEO Chuck Robbins has stated that this will drive traffic “beyond the peaks of current chatbot interaction” and keep network levels “consistently high”.
- Edge-based AI: A third wave of “physical AI” will require more computing and networking at the edge of the network to accommodate specialized use cases like industrial IoT.
“As we move towards agentic AI and the demand for inferencing expands to the enterprise and end user networking environments, traffic on the network will reach unprecedented levels,” Cisco CEO Chuck Robbins said on the company’s recent earnings call. “Network traffic will not only increase beyond the peaks of current chatbot interaction, but will remain consistently high with agents in constant interaction.”
2. Nokia recently predicted that both direct and indirect AI traffic on mobile networks will grow at a faster pace than regular, non-AI traffic.
- Direct AI traffic: This is generated by users or systems directly interacting with AI services and applications. Consumer examples: Using generative AI tools, interacting with AI-powered gaming, or experiencing extended reality (XR) environments. Enterprise examples: Employing predictive maintenance, autonomous operations, video and image analytics, or enhanced customer interactions.
- Indirect AI traffic: This occurs when AI algorithms are used to influence user engagement with existing services, thereby increasing overall traffic. Examples: AI-driven personalized recommendations for video content on social media, streaming platforms, and online marketplaces, which can lead to longer user sessions and higher bandwidth consumption.
The Finland based network equipment vendor warned that the AI wave could bring “a potential surge in uplink data traffic that could overwhelm our current network infrastructure if we’re not prepared,” noting that the rise of hybrid on-device and cloud tools will require much more than the 5-15 Mbps uplink available on today’s networks. Nokia’s Global Network Traffic 2030 report forecasts that overall traffic could grow by 5 to 9 times current levels by 2033. All told, Nokia said AI traffic is expected to hit 1088 exabytes (EB) per month by 2033. That means overall traffic will grow 5x in a best case scenario and 9x in a worse case.
To manage this anticipated traffic surge, Nokia advocates for radical changes to existing network infrastructure.
- Cognitive networks: The company states that networks must become “cognitive,” leveraging AI and machine learning (ML) to handle the growing data demand.
- Network-as-Code: As part of its Technology Strategy 2030, Nokia promotes a framework for more flexible and scalable networks that leverage AI and APIs.
- 6G preparation: Nokia Bell Labs is already conducting research and field tests to prepare for 6G networks around 2030, with a focus on delivering the capacity needed for AI and other emerging technologies.
- Optimizing the broadband edge: The company also highlights the need to empower the broadband network edge to handle the demands of AI applications, which require low latency and high reliability.
Nokia’s Global Network Traffic 2030 report didn’t mention agentic AI, which are artificial intelligence systems designed to autonomously perceive, reason, and act in their environment to achieve complex goals with less human oversight. Unlike generative AI, which focuses on creating content, agentic AI specializes in workflow automation and independent problem-solving by making decisions, adapting plans, and executing tasks over extended periods to meet long-term objectives.
3. Ericsson did point to traffic increases stemming from the use of AI-based assistants in its 2024 Mobility Report. In particular, it predicted the majority of traffic would be related to the use of consumer video AI assistants, rather than text-based applications and – outside the consumer realm – forecast increased traffic from “AI agents interacting with drones and droids. Accelerated consumer uptake of GenAI will cause a steady increase of traffic in addition to the baseline increase,” Ericsson said of its traffic growth scenario.
…………………………………………………………………………………………………………………………………………………………………………………..
Dissenting Views:
1. UK Disruptive Analysis Founder Dean Bubley isn’t a proponent of huge AI traffic growth. “Many in the telecom industry and vendor community are trying to talk up AI as driving future access network traffic and therefore demand for investment, spectrum etc., but there is no evidence of this at present,” he told Fierce Network.
Bubley argues that AI agents won’t really create much traffic on access networks to homes or businesses. Instead, he said, they will drive traffic “inside corporate networks, and inside and between data centers on backbone networks and inside the cloud. “There might be a bit more uplink traffic if video/images are sent to the cloud for AI purposes, but again that’s hypothetical,” he said.
2. In a LinkedIn post, Ookla analyst Mike Dano said he was a bit suspicious about “Cisco predicting a big jump in network traffic due to AI agents constantly wandering around the Internet and doing things.” While almost all of the comments agreed with Dano, it still is an open question whether the AI traffic Armageddon will actually materialize.
……………………………………………………………………………………………………………………………………………………………………………………….
References:
RtBrick survey: Telco leaders warn AI and streaming traffic to “crack networks” by 2030
https://www.fierce-network.com/cloud/will-ai-agents-really-raise-network-traffic-baseline
Mulit-vendor Open RAN stalls as Echostar/Dish shuts down it’s 5G network leaving Mavenir in the lurch
Last week’s announcement that Echostar/ Dish Network will sell $23 billion worth of spectrum licenses to AT&T was very bad news for Mavenir. As a result of that deal, Dish Network’s 5G Open RAN network, running partly on Mavenir’s software, is to be decommissioned. Dish Network had been constructing a fourth nationwide U.S. mobile network with new Open RAN suppliers – one of the only true multi-vendor Open RANs worldwide.
Credit: Kristoffer Tripplaar/Alamy Stock Photo
Echostar’s decision to shut down its 5G network marks a very sad end for the world’s largest multivendor open RAN and will have ramifications for the entire industry. “If you look at all the initiatives, what the US government did or in general, they are the only ones who actually spent a good chunk of money to really support open RAN architecture,” said Pardeep Kohli, the CEO of Mavenir, one of the vendors involved in the Dish Network project. “So now the question is where do you go from here?”
As part of its original set of updates on 5G network plans, Dish revealed it would host its 5G core – the part that will survive the spectrum sale – in the public cloud of AWS. And the hyperscaler’s data facilities have also been used for RAN software from Mavenir installed on servers known as central units.
Open RAN enters is in the fronthaul interface between Mavenir’s DU software and radios provided by Japan’s Fujitsu. Its ability to connect its software to another company’s radios validates Mavenir’s claims to be an open RAN vendor, says Kohli. While other suppliers boast compatibility with open RAN specifications, commercial deployments pairing vendors over this interface remain rare.
Mavenir has evidently been frustrated by the continued dominance of Huawei, Ericsson and Nokia, whose combined RAN market share grew from 75.1% in 2023 to 77.5% last year, according to research from Omdia, an Informa company. Dish Network alone would not have made a sufficient difference for Mavenir and other open RAN players, according to Kohli. “It helped us come this far,” he said. “Now it’s up to how far other people want to take it.” A retreat from open RAN would, he thinks, be a “bad outcome for all the western operators,” leaving them dependent on a Nordic duopoly in countries where Chinese vendors are now banned.
“If they (telcos) don’t support it (multi-vendor OpenRAN), and other people are not supporting it, we are back to a Chinese world and a non-Chinese world,” he said. “In the non-Chinese world, you have Ericsson and Nokia, and in the Chinese world, it’s Huawei and ZTE. And that’s going to be a pretty bad outcome if that’s where it ends up.”
…………………………………………………………………………………………………………………………………………………………………
Open RAN x-U.S.:
Outside the U.S., the situation is no better for OpenRAN. Only Japan’s Rakuten and Germany’s 1&1 have attempted to build a “greenfield” Open RAN from scratch. As well as reporting billions of dollars in losses on network deployment, Rakuten has struggled to attract customers. It owns the RAN software it has deployed but counts only 1&1 as a significant customer. And Rakuten’s original 4G rollout was not based on the industry’s open RAN specifications, according to critics. “They were not pure,” said Mavenir’s Kohli.
Plagued by delays and other problems, 1&1’s rollout has been a further bad advert for Open RAN. For the greenfield operators, the issue is not the maturity of open RAN technology. Rather, it is the investment and effort needed to build any kind of new nationwide telecom network in a country that already has infrastructure options. And the biggest brownfield operators, despite professing support for open RAN, have not backed any of the the new entrants.
RAN Market Concentration:
- Stefan Pongratz, an analyst with Dell’Oro, found that five of six regions he tracks are today classed as “highly concentrated,” with an HHI score of more than 2,500. “This suggests that the supplier diversity element of the open RAN vision is fading,” wrote Pongratz in a recent blog.
- A study from Omdia (owned by Informa), shows the combined RAN market share of Huawei, Ericsson and Nokia grew from 75.1% in 2023 to 77.5% last year. The only significant alternative to the European and Chinese vendors is Samsung, and its market share has shrunk from 6.1% to 4.8% over this period.
Concentration would seem to be especially high in the U.S., where Ericsson now boasts a RAN market share of more than 50% and generates about 44% of its sales (the revenue contribution of India, Ericsson’s second-biggest market, was just 4% for the recent second quarter). That’s partly because smaller regional operators previously ordered to replace Huawei in their networks spent a chunk of the government’s “rip and replace” funds on Ericsson rather than open RAN, says Kohli. Ironically, though, Ericsson owes much of the recent growth in its U.S. market share to what has been sold as an open RAN single vendor deal with AT&T [1.]. Under that contract, it is replacing Nokia at a third of AT&T’s sites, having already been the supplier for the other two thirds.
Note 1. In December 2023, AT&T awarded Ericsson a multi-year, $14 billion Open RAN contract to serve as the foundation for its open network deployment, with a goal of having 70% of its wireless traffic on open platforms by late 2026. That large, single-vendor award for the core infrastructure was criticized for potentially undermining the goal of Open RAN which was to encourage competition among multiple network equipment and software providers. AT&T’s claim of a mulit-vendor network turned out to be just a smokescreen. Fujitsu/1Finity supplied third-party radios used in AT&T’s first Open RAN call with Ericsson.
Indeed, AT&T’s open RAN claims have been difficult to take seriously, especially since it identified Mavenir as a third supplier of radio units, behind Ericsson and Japan’s Fujitsu, just a few months before Mavenir quit the radio unit market. Mavenir stopped manufacturing and distributing Open RAN radios in June 2025 as part of a financial restructuring and a shift to a software-focused business model.
…………………………………………………………………………………………………………………….
Arguably, Kohli describes Echostar/ Dish Network as the only U.S. player that was spending “a good chunk of money to really support open RAN architecture.”
Ultimately, he thinks the big U.S. telcos may come to regret their heavier reliance on the RAN gear giants. “It may look great for AT&T and Verizon today, but they’ll be funding this whole thing as a proprietary solution going forward because, really, there’s no incentive for anybody else to come in,” he said.
…………………………………………………………………………………………………………………….
References:
https://www.lightreading.com/open-ran/echostar-rout-leaves-its-open-ran-vendors-high-and-dry
AT&T to to buy spectrum Licenses from EchoStar for $23 billion
AT&T to deploy Fujitsu and Mavenir radio’s in crowded urban areas
Dell’Oro Group: RAN Market Grows Outside of China in 2Q 2025
Mavenir and NEC deploy Massive MIMO on Orange’s 5G SA network in France
Spark New Zealand completes 5G SA core network trials with AWS and Mavenir software
Mavenir at MWC 2022: Nokia and Ericsson are not serious OpenRAN vendors
Ericsson expresses concerns about O-RAN Alliance and Open RAN performance vs. costs
Nokia and Mavenir to build 4G/5G public and private network for FSG in Australia
Hyperscaler design of networking equipment with ODM partners
Networking equipment for hyperscalers like Google, Amazon, Microsoft, Oracle, Meta, and others is a mix of in‑house engineering and partnerships with specialized vendors. These companies operate at such massive scale to design their own switches, routers, and interconnects — but rely on Original Design Manufacturers (ODMs) and network silicon providers to build them.
In‑House Networking Design:
Hyperscalers have dedicated hardware teams that create custom network gear to meet their unique performance, latency, and power‑efficiency needs.
- Google – Designs its own Jupiter and Andromeda data center network fabrics, plus custom top‑of‑rack (ToR) and spine switches. Uses merchant silicon from Broadcom, Intel (Barefoot Tofino), and others, but with Google‑built control planes and software.
- Amazon (AWS) – Builds custom switches and routers for its Scalable Reliable Datagram (SRD) and Elastic Fabric Adapter (EFA) HPC networks. Uses in‑house firmware and network operating systems, often on ODM‑built hardware.
- Microsoft (Azure) – Designs OCP‑compliant switches (e.g., SONiC network OS) and contributes to the Open Compute Project. Uses merchant silicon from Broadcom, Marvell, and Mellanox/NVIDIA.
- Oracle Cloud Infrastructure (OCI) – Designs its own high‑performance RDMA‑enabled network for HPC and AI workloads, with custom switches built by ODM partners.
- Meta – Designs Wedge, Backpack, and Minipack switches under OCP, manufactured by ODMs.
Manufacturing & ODM Partners:
While the hyperscaler’s network equipment designs are proprietary, the physical manufacturing is typically outsourced to ODMs who specialize in hyperscale networking gear:
| ODM / OEM | Builds for | Notes |
|---|---|---|
| Quanta Cloud Technology (QCT) | AWS, Azure, Oracle, Meta | Custom ToR/spine switches, OCP gear |
| WiWynn | Microsoft, Meta | OCP‑compliant switches and racks |
| Celestica | Multiple hyperscalers | High‑end switches, optical modules |
| Accton / Edgecore | Google, Meta, others | White‑box switches for OCP |
| Foxconn / Hon Hai | AWS, Google | Large‑scale manufacturing |
| Delta Networks | Multiple CSPs | Optical and Ethernet gear |
Network Silicon & Optics Suppliers:
Even though most hyperscalers design the chassis and racks, they often use merchant silicon and optics from:
- Broadcom – Tomahawk, Trident, Jericho switch ASICs
- Marvell – Prestera switch chips, OCTEON DPUs
- NVIDIA (Mellanox acquisition) – Spectrum Ethernet, InfiniBand for AI/HPC
- Intel (Barefoot acquisition) – Tofino programmable switch ASICs
- Cisco Silicon One – Used selectively in hyperscale builds
- Coherent optics & transceivers – From II‑VI (Coherent), Lumentum, InnoLight, etc.
Hyperscaler Networking Supply Chain Map:
| Layer | Key Players | Role | Example Hyperscaler Relationships |
|---|---|---|---|
| Network Silicon (ASICs / DPUs) | Broadcom (Tomahawk, Jericho), Marvell (Prestera, OCTEON), NVIDIA/Mellanox (Spectrum, InfiniBand), Intel (Barefoot Tofino), Cisco (Silicon One) | Core packet switching, programmability, congestion control | Google (Broadcom, Intel), AWS (Broadcom, Marvell), Microsoft (Broadcom, Marvell, NVIDIA), Oracle (Broadcom, NVIDIA) |
| Optics & Interconnects | Coherent (II‑VI), Lumentum, InnoLight, Source Photonics, Broadcom (optical PHYs) | 400G/800G transceivers, co‑packaged optics, DWDM modules | All hyperscalers source from multiple vendors for redundancy |
| ODM / Manufacturing | Quanta Cloud Technology (QCT), WiWynn, Celestica, Accton/Edgecore, Foxconn, Delta Networks | Build hyperscaler‑designed switches, routers, and chassis | AWS (QCT, Foxconn), Google (Accton, QCT), Microsoft (WiWynn, Celestica), Meta (Accton, WiWynn), Oracle (QCT, Celestica) |
| Network OS & Control Plane | In‑house NOS (Google proprietary, AWS custom OS, Microsoft SONiC, Oracle custom), OCP software | Routing, telemetry, automation, SDN control | Google (Jupiter fabric OS), AWS (custom SRD/EFA stack), Microsoft (SONiC), Oracle (OCI NOS) |
| Integration & Deployment | Hyperscaler internal engineering teams | Rack integration, cabling, fabric topology, automation pipelines | All hyperscalers do this in‑house for security and scale |
Design Flow:
- Chip Vendors → supply merchant silicon to ODMs or directly to hyperscaler design teams.
- Hyperscaler Hardware Teams → design chassis, PCB layouts, thermal systems, and specify optics.
- ODMs → manufacture to spec, often in Asia, with hyperscaler QA oversight.
- Optics Vendors → deliver transceivers and cables, often qualified by hyperscaler labs.
- In‑House NOS → loaded onto hardware, integrated into hyperscaler’s SDN fabric.
- Deployment → rolled out in data centers globally, often in multi‑tier Clos or AI‑optimized topologies.
Major Trends:
- Disaggregation – Hyperscalers separate hardware from software, running their own Network Operating System (NOS), (e.g., SONiC, Google’s proprietary OS) on ODM‑built “white‑box” or “bare metal” switches.
- AI‑Optimized Fabrics – New designs focus on ultra‑low latency, congestion control, and massive east‑west bandwidth for GPU clusters.
- Optical Integration – Co‑packaged optics and 800G+ transceivers are becoming standard for AI and HPC workloads.
- AI Cluster Networking – NVIDIA InfiniBand and 800G Ethernet fabrics are now common for GPU pods.
- Co‑Packaged Optics – Moving optics closer to the ASIC to reduce power and latency.
- Open Compute Project Influence – Many designs are OCP‑compliant but with proprietary tweaks.
- Multi‑Vendor Strategy – Hyperscalers dual‑source ASICs and optics to avoid supply chain risk.
References:
How it works: hyperscaler compute server in house design process with ODM partners
Cloud‑resident high performance compute servers used by hyperscale cloud providers like Amazon Web Services (AWS), Google Cloud Platform (GCP), Microsoft Azure, Oracle Cloud Infrastructure (OCI), Meta, and others use a mix of custom in‑house designs and ODM (Original Design Manufacturer) ‑ built hardware.
In‑House Design Teams:
- Amazon (AWS) – Designs its own Nitro System–based servers, including custom motherboards, networking cards, and security chips. AWS also develops Graviton (Arm‑based) and Trainium/Inferentia (AI) processors. HPC instances use Elastic Fabric Adapter (EFA) for low‑latency interconnects.
- Google (GCP) – Builds custom server boards and racks for its data centers, plus TPUs (Tensor Processing Units) for AI workloads. GCP builds custom HPC server boards and racks, plus TPUs for AI workloads. It uses high‑speed interconnects like Google’s Jupiter network for HPC clusters.
- Microsoft Azure – Designs Azure‑optimized servers and AI accelerators, often in collaboration with partners, and contributes designs to the Open Compute Project (OCP). It integrates InfiniBand and/or 400 Gbps Ethernet for HPC interconnects.
- Oracle – Designs bare‑metal HPC shapes with AMD EPYC, Intel Xeon, and NVIDIA GPUs, plus RDMA cluster networking for microsecond latency.
- Meta – Designs its compute servers, especially for AI workloads, by working closely with ODM partners like Quanta Computer, Wiwynn, and Foxconn.
Manufacturing Partners (ODMs/OEMs):
While the hyperscaler compute server designs are proprietary, the physical manufacturing is typically outsourced to Original Design Manufacturers (ODMs) who specialize in hyperscale data center gear as per these tables:
| ODM / OEM | Known for | Cloud Customers |
|---|---|---|
| Quanta Cloud Technology (QCT) | Custom rack servers, HPC nodes | AWS, Azure, Oracle |
| WiWynn | OCP‑compliant HPC servers | Microsoft, Meta |
| Inventec | HPC and AI‑optimized servers | AWS, GCP |
| Foxconn / Hon Hai | Large‑scale server manufacturing | Google, AWS |
| Celestica | HPC and networking gear | Multiple hyperscalers |
| Supermicro | GPU‑dense HPC systems | AWS, Oracle, Azure |
| ODM / OEM | Role in Hyperscale Cloud |
|---|---|
| Quanta Cloud Technology (QCT) | Major supplier for AWS, Azure, and others; builds custom rack servers and storage nodes. |
| WiWynn | Spun off from Wistron; manufactures OCP‑compliant servers for Microsoft and Facebook/Meta. |
| Inventec | Supplies compute and storage servers for AWS and other CSPs. |
| Foxconn / Hon Hai | Builds cloud server hardware for multiple providers, including Google. |
| Delta / Celestica | Provides specialized server and networking gear for hyperscale data centers. |
| Supermicro | Supplies both standard and custom AI‑optimized servers to cloud and enterprise customers. |
The global server market expected to reach $380 billion by 2028. Image credit: Alamy
……………………………………………………………………………………………………………………………………………………………..
Here’s a supply chain relationship map for cloud‑resident high‑performance compute (HPC) servers used by the major hyperscalers:
Hyperscale HPC Server Design & Manufacturing Landscape:
| Cloud Provider | In‑House Design Focus | Key Manufacturing / ODM Partners | Notable HPC Hardware Features |
|---|---|---|---|
| Amazon Web Services (AWS) | Custom Nitro boards, Graviton CPUs, Trainium/Inferentia AI chips, EFA networking | Quanta Cloud Technology (QCT), Inventec, Foxconn | Arm‑based HPC nodes, GPU clusters (NVIDIA H100/A100), ultra‑low‑latency RDMA |
| Google Cloud Platform (GCP) | Custom server boards, TPU accelerators, Jupiter network fabric | Quanta, Inventec, Foxconn | TPU pods, GPU supernodes, liquid‑cooled racks |
| Microsoft Azure | OCP‑compliant HPC designs, Maia AI chip, Cobalt CPU, InfiniBand networking | WiWynn, QCT, Celestica | Cray‑based HPC clusters, GPU/FPGA acceleration |
| Oracle Cloud Infrastructure (OCI) | Bare‑metal HPC shapes, RDMA cluster networking | QCT, Supermicro | AMD EPYC/Intel Xeon nodes, NVIDIA GPU dense racks |
| Meta (for AI/HPC research) | OCP‑based AI/HPC servers | WiWynn, QCT | AI Research SuperCluster, liquid cooling |
| Alibaba Cloud / Tencent Cloud | Custom AI/HPC boards, Arm CPUs | Inspur, Sugon, QCT | GPU/FPGA acceleration, high‑bandwidth fabrics |
Meta’s ODM Collaboration Model:
- Quanta Computer: Meta has partnered with Quanta for final assembly of its next-gen AI servers. Quanta is responsible for building up to 6,000 racks of the Santa Barbara servers, which feature advanced cooling and power delivery systems.
- Wiwynn & Foxconn: These ODMs also play key roles in Meta’s infrastructure. Wiwynn reportedly earns more than half its revenue from Meta, while Foxconn handles system assembly for NVIDIA’s NVL 72 servers, which Meta may also utilize.
- Broadcom Partnership: For chip supply, Meta collaborates with Broadcom to integrate custom ASICs into its server designs.
Hyperscaler/ODM Collaboration Process:
- Design Phase – Hyperscalers’ hardware teams define the architecture: CPU/GPU choice, interconnect, cooling, power density.
- ODM Manufacturing – Partners like Quanta, WiWynn, Inventec, Foxconn, Celestica, and Supermicro build the servers to spec.
- Integration & Deployment – Systems are tested, integrated into racks, and deployed in hyperscale data centers.
- Optimization – Providers fine‑tune firmware, drivers, and orchestration for HPC workloads (e.g., CFD, genomics, AI training).
Industry Trends:
- Open Compute Project (OCP) – Many designs are shared in the OCP community, allowing ODMs to build interoperable, cost‑optimized hardware at scale. Open Compute Project designs speed up deployment and interoperability.
- Vertical Integration – Hyperscalers increasingly design custom silicon (e.g., AWS Graviton, Google TPU, Microsoft Maia AI chip) to optimize performance and reduce dependency on third‑party CPUs/GPUs.
- AI‑Optimized Racks – New designs focus on high‑density GPU clusters, liquid cooling, and ultra‑low‑latency networking for AI workloads.
- Vertical integration: More custom silicon to optimize performance and cost. See Specialized HPC components below.
- Liquid cooling: Increasingly common for dense GPU/CPU HPC racks.
Specialized HPC Components:
- CPUs – AMD EPYC, Intel Xeon Scalable, AWS Graviton (Arm), custom Google CPUs.
- GPUs / Accelerators – NVIDIA H100/A100, AMD Instinct, Google TPU, AWS Trainium.
- Networking – Mellanox/NVIDIA InfiniBand, AWS EFA, Oracle RDMA cluster networking.
- Storage – Parallel file systems like Lustre, BeeGFS, IBM Spectrum Scale for HPC workloads
References:
ODM Sales Soar as Hyperscalers and Cloud Providers Go Direct
The future of US hyperscale data centers | McKinsey
100MW+ Wholesale Colocation Deals: Inside the Hyperscaler Surge
Hyperscaler design of networking equipment with ODM partners – IEEE ComSoc Technology Blog
Liquid Dreams: The Rise of Immersion Cooling and Underwater Data Centers
