AST SpaceMobile booked a big year-on-year rise in revenue in Q2, driven by gateway deliveries and milestones tied to US government contracts as the direct-to-device (D2D) satellite operator continues to scale ahead of a planned beta service launch later this year.

The company stated it remains on track to meet its full-year 2026 revenue guidance of $150 million to $200 million on the back of additional contract awards from the US government.

Revenue backlog grew to approximately $1.3 billion in aggregate contracts with commercial partners and government.

Chair and CEO Abel Avellan said the company’s technology platform, spectrum strategy and manufacturing capabilities position it to deliver D2D connectivity for commercial and government customers globally, pointing to the growth of its constellation to 13 spacecraft in orbit

On its earnings call, president Scott Wisniewski highlighted growing momentum with US government customers who he said view the company’s technology as unique, strategic, innovative and flexible for communications and non-communications uses.

He said AST SpaceMobile previously flagged a trend of small development contracts evolving into larger awards, eventually maturing into full programmes of record.

“Today, you can see that trend as we are announcing three new contract awards with funded near-term value of over $100 million in total expected during 2026 and 2027.”

Further public details on the awards are expected soon, but Wisniewski described them as capabilities developed in partnership with the US Department of War over several years which employ the company’s technology to address a significant strategic need.

Trends
AST SpaceMobile booked revenue of $31.5 million compared with $1.2 million in Q2 2025.

Net loss attributable to common stockholders grew from $99.4 million to $230.9 million.

Trends in the opening half followed suit, with revenue of $46.3 million compared with $1.9 million, and attributable net loss up from $145.1 million to $421.9 million.

The post AST SpaceMobile revenue rockets appeared first on Mobile World Live.

Nvidia inked strategic partnerships with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to establish AI compute infrastructure financing platforms to mobilise more than $500 billion of third-party capital to build related infrastructure.

The chipmaker stated memorandums of understanding would help create the first compute financing platforms of their kind at global scale, intended to fund frontier AI laboratories, clouds and enterprises.

It noted the platforms are designed to create dedicated pools of capital at significant scale and attractive rates for its customers, positioning Nvidia compute and full-stack AI infrastructure as an investable asset class.

“We are bringing the world’s leading long-term capital providers together to independently underwrite AI infrastructure,” Nvidia CEO Jensen Huang said.

“These financing platforms will help customers access scarce compute at scale and build the DSX AI factories that will power every industry and country”.

Nvidia described its compute as an investable asset offering the lowest token cost, highest revenue and longest life, alongside a rich ecosystem of buyers.

The arrangement builds on Nvidia’s broader push to raise capital for it and its customers to continue producing the chips, power systems and data centres underpinning the AI boom.

Nvidia GPUs underpin most leading US AI models available today and it supplies relevant chips, infrastructure, and software to a wide range of partners.

Money matters
It faced scrutiny in recent months over circular financing arrangements in which the company provides backing to help AI partners raise debt, in turn boosting Nvidia’s own revenue.

Bringing outside private capital providers onboard may help address some of the criticism since it spreads the financing load beyond Nvidia itself.

In recent years, private capital groups including Apollo and Blackstone have structured AI deals to help companies including Anthropic finance heavy spending on chips and data centres.

Financial Times reported Morgan Stanley predicts large cloud computing players would spend $3.5 trillion on AI infrastructure over the next two years.

The post Investors put $500B into Nvidia AI infrastructure plan appeared first on Mobile World Live.

Meta Platforms CEO Mark Zuckerberg (pictured) argued open, decentralised access to super-intelligence is essential for ensuring global economic prosperity and personal safety.

In a lengthy blog, the executive rejected doom-centric narratives surrounding advanced AI technologies.

Rather than centralising powerful models within a select group of corporations or government institutions, Zuckerberg advocated for a balance-of-power approach, which puts personal superintelligence directly into the hands of individuals.

On the same day the post was published, Meta Superintelligence Labs released the open weights for Muse Glimmer, a 30-billion-parameter dense multimodal model under an Apache 2.0 license on Hugging Face.

The model is distilled from Muse Spark 1.2 and engineered to run locally on a single consumer GPU or Mac to execute agentic workflows such as autonomous coding and file management without cloud dependency.

Meta also plans to publish open weights for its Muse Spark 1.2 model in the coming weeks, reversing a brief closed-weight stance, while teasing a next generation frontier model known internally as Watermelon.

The release and blog came amid heightened scrutiny over closed-source AI labs following recent incidents where models from OpenAI and Anthropic broke containment during evaluation testing.

Zuckerberg positioned open-source distribution as a key countermeasure to security risks. He stated broad developer access allows system defences to be patched faster than in centralised environments.

He also defended AI distillation practices by claiming learning from observed model outputs is essential for keeping US open-source technology competitive against foreign rivals like DeepSeek.

Central to Meta’s broader strategy is the deployment of personal AI agents designed to assist users across daily tasks, business operations and learning.

To support this vision Meta unveiled a $1 billion community fund for US cities hosting its data centres to address local power, water and infrastructure needs, while committing to train skilled trade workers through its Community Compact and America’s Workforce Academy initiatives.

The post Meta CEO makes latest AI for everyone pitch appeared first on Mobile World Live.

Intel proposed a massive $15 billion underwritten public offering of common stock to bolster its balance sheet and aggressively fund its expansion in the high-stakes AI and semiconductor markets.

Financial underwriters include J.P. Morgan Securities, Goldman Sachs, Morgan Stanley and Citigroup Global Markets. Intel stated the companies would have the option for 30-days to purchase up to an additional $2.3 billion in shares, a strategy to secure long-term competitiveness.

“Progress in emerging areas including physical AI, purpose-built silicon, advanced packaging and external wafers represent significant growth opportunities for Intel,” the company stated.

The additional capital targets soaring global demand for technology, as customer investments in AI compute power and physical iterations continue to grow rapidly.

Intel stated net proceeds would cover “general corporate purposes” including working capital and capital expenditures. The company expects the offering to help preserve its investment-grade credit rating during heavy infrastructure investment.

The post Intel launches $15B stock sale to fund AI ambitions appeared first on Mobile World Live.

Anthropic formed a strategic partnership with Macquarie Asset Management and Singaporean wealth fund GIC to build data centres which the AI company would lease under long-term arrangements.

The Theseus Infrastructure platform is to be a vehicle to develop AI computing sites with an initial focus on the US.

Macquarie Asset Management and GIC pledged to fund the bulk of the equity for each project, while Anthropic committed to covering any consumer electricity price hikes stemming from the facilities.

The asset management company stated sites would require a major investment and highlighted anticipated boosts for employment covering building and operation.

GIC is a repeat Anthropic investor.

It jointly led a $30 billion funding round in February and backed a $1.5 billion enterprise services venture Anthropic launched in May.

In 2025, Anthropic stated it would spend $50 billion on custom data centres across several US locations, including the states of Texas and New York.

The company is also preparing for a public listing.

Bloomberg reported Anthropic has made relatively few high-profile data centre partnerships compared with rival OpenAI, which launched its Stargate infrastructure joint venture in 2025.

The post Anthropic, Macquarie, GIC to build AI data centres appeared first on Mobile World Live.

OpenAI temporarily paused certain internal activities surrounding its upcoming model after the system demonstrated unprecedented cybersecurity and agentic coding capabilities.

The ChatGPT maker disclosed preliminary evaluations showed the Astra model is significantly adept at cyber tasks, leading company officials to admit they cannot rule out it becoming a critical security threat threshold.

Under OpenAI’s internal Preparedness Framework, a critical threshold is defined by a model’s ability to autonomously discover and execute zero-day exploits across hardened real-world systems without human intervention, or independently carry out complex attack strategies.

While evaluations remain ongoing alongside third-party experts, the initial findings prompted immediate containment protocols.

OpenAI is moving Astra into isolated testing environments equipped with sandboxed execution, restricted tool and network access, enhanced model weight encryption and full chain-of-thought monitoring.

It stated it paused related activities which do not currently meet the safety controls.

OpenAI’s move follows a recent series of incidents where autonomous AI agents breached security boundaries during testing.

Within the past month, OpenAI, Anthropic and Meta Platforms have all publicly acknowledged inadvertent system intrusions during evaluation phases.

OpenAI plans to partner with government agencies and AI safety institutes to evaluate Astra’s defensive and attacking potential before considering public release.

It emphasised its goal is to equip cyber defence players with high-capability models to patch vulnerabilities before malicious actors can exploit them.

.

The post OpenAI halts Astra model testing appeared first on Mobile World Live.

Intel appointed Marvell executive Dean Jarnac (pictured) as EVP and chief sales officer as the chipmaker seeks to bolster customer engagement and maintain momentum from recent high-profile foundry wins.

The incoming executive will lead Intel’s global sales organisation across its client, data centre, AI, networking and ASIC portfolios.

Jarnac brings over three decades of semiconductor industry experience. In addition to Marvell, he previously held senior leadership roles at Broadcom and AMD.

He will assume the role in September and report directly to CEO Lip-Bu Tan. Jarnac will take over from Greg Ernst, who is departing Intel after 27 years with the company.

Announcing the appointment, Tan highlighted customer focus and operational execution are central to Intel’s growth strategy. He cited Jarnac’s track record of building high-performance sales teams as critical to deepening client partnerships.

The change follows significant progress in Intel’s foundry turnaround strategy. Earlier this year, Google placed an order with Intel to manufacture more than 3 million custom tensor processing units (TPUs) in 2028.

In April, Intel revealed a plan to join Elon Musk’s Terafab AI chip project to build processors for the billionaire’s orbital data centres and humanoid robots.

Apple and Intel reportedly reached a preliminary agreement in May for the latter to manufacture some of the chips for iPhones, iPads and other devices.

The post Intel names new chief sales officer appeared first on Mobile World Live.

MobileX and Boost Mobile founder Peter Adderton (pictured) rebuffed the reported view of T-Mobile US CEO Srini Gopalan that Starlink poses little threat to mobile operators.

Financial Times (FT) reported Gopalan is struggling to see what consumer problem satellite services solve by entering the mobile market.

“You’ve got to ask the question of ‘What is the gap in the market that they’re looking to serve?'” he reportedly said, adding a broader network is “clearly not” what would draw customers to a new mobile service.

In a post on LinkedIn, Adderton said Gopalan’s position misses the point entirely.

“I always find it funny when wireless executives ask, ‘Why would a customer switch? It’s clearly not because they have a broader cellular network,'” he said. “That’s exactly the point.”

He argued Starlink is not trying to out-build T-Mobile’s network, but to offer something terrestrial operators cannot easily replicate: a single service combining terrestrial and satellite connectivity, spanning cities, rural areas, highways, mountains and water, with plans to eventually extend worldwide.

Adderton drew a comparison to Uber, arguing consumers did not choose the service because it built better roads, but because it created a better experience on infrastructure which already existed.

He explained the mistake operators make is assuming customers are shopping for another mobile network, when they may instead be shopping for connectivity which “just works”.

Adderton also took aim at the broader response from operator executives to Starlink’s mobile ambitions, suggesting a coordinated messaging effort.

“Seems every MNO CEO in the JV got the same script and, credit where credit’s due, they’re sticking to it beautifully,” he wrote on X. “Their PR teams must be very proud. Not a rogue talking point in sight.”

Plans
Gopalan’s reported comments were made before Elon Musk confirmed SpaceX plans to build a US mobile service.

On SpaceX’s Q2 earnings call, Musk said Starlink’s terrestrial mobile network would likely offer better connectivity and higher bandwidth than what is currently available from mobile operators.

SpaceX president Gwynne Shotwell said she expects to draw a significant number of customers from AT&T, Verizon, and T-Mobile, with Starlink Mobile satellites launching in 2027 and services due to commence by the year-end.

Gopalan told FT satellite technology would remain “complementary” to existing networks rather than a replacement.

T-Mobile is one of several operators which holds satellite access deals with SpaceX to supplement rural coverage.

FT reported Gopalan also addressed T-Mobile’s growth strategy, stating it aims to more than double its broadband customer base to 20 million within four years and grow its rural market share beyond its current 24%.

The post MobileX founder counters T-Mobile CEO on Starlink threat appeared first on Mobile World Live.

The US Federal Communications Commission (FCC) voted to explore ways to deliver direct satellite connectivity to unlicensed devices using 200MHz of related spectrum to advance services requiring handoff between terrestrial and non-terrestrial links.

It stated the move could give consumers an “additional option for reliable connectivity”, as investment in direct-to-device (D2D) services continues to grow.

The Notice of Proposed Rulemaking (NPRM) examines potential reforms to enable devices operating in certain part 15 bands communicate with FCC-authorised satellites.

Part 15 rules govern unlicensed wireless equipment widely used by businesses and consumers. The FCC stated the move is intended to expand opportunities for Wi-Fi and applications including IoT which “benefit from more robust wireless connectivity”.

It also proposed clarifying part 15-authorised equipment may be used within and between authorised spacecraft.

The FCC intends to explore further scenarios “in which part 15 devices might” safely and effectively operate in space,

Money maker
FCC chair Brendan Carr said the move builds on 18-months of regulatory steps to enable D2D services, including freeing up spectrum, providing certainty around buildout milestones, and clarifying exclusivity and legal rights to encourage investment.

He said the measures had helped generate tens of billions of dollars in D2D deal flow.

“For the first time in FCC history, we are exploring whether D2D services are technically possible in more than 200 MHz of unlicensed spectrum,” he said.

“These frequencies could yield more capacity to complement D2D services and encourage experimentation.”

He added unlicensed spectrum already underpins a wide range of household devices and innovations, and the proposal would seek to ensure the bands continue to support existing services.

The proposal targets the 902MHz to 928MHz, 2400MHz to 2483.5MHz and 5725MHz to 5850MHz bands traditionally used for Wi-Fi and Bluetooth which already permit unlicensed devices to operate at the relatively higher power levels needed for satellite communication.

Along with seeking comments, the FCC is asking whether part 15 devices communicating with satellites should require an additional part 25 earth station certification.

The post FCC eyes unlicensed spectrum for D2D appeared first on Mobile World Live.

Tampa Electric Company (TECO) selected Ericsson and OneLayer to deploy, manage and secure its growing private LTE network, with Burns & McDonnell providing integration and engineering expertise.

Ericsson stated TECO is among the first investor-owned utilities in the US to deploy the private network technology for operational use.

The network uses the Swedish vendor’s kit and supports connectivity for operational devices across TECO’s delivery infrastructure. It will expand as more connected devices come online in the coming years.

OneLayer’s Bridge platform provides SIM lifecycle management and device onboarding. It also provides real time visibility into connected devices across TECO’s operational infrastructure.

Burns & McDonnell is responsible for aligning the deployment with existing systems and operational requirements.

Following the deployment, TECO’s operations teams will be able to onboard and provision cellular-connected devices through automated workflow, which Ericsson noted cuts down the manual effort needed to bring new assets onto the network.

Teams also gain visibility into devices behind cellular routers and customer premises equipment, which were previously difficult to monitor. The system allows the tracking of device status location, and connectivity in real time across distributed field sites.

TECO plans to deploy tens of thousands of devices over the next five years with several thousand targeted in the first phase.

Ericsson stated the approach gives TECO a private, high-reliability broadband platform. It supports grid protection, field workforce communications and sensor connectivity. It also provides a path to evolve to 5G as requirements grow.

Dana Jaber, head of mission critical networks at Ericsson Americas, stated the deployment sets a new benchmark “for how utilities can scale secure, high-reliability private cellular networks to support tens of thousands of devices and future 5G requirements”.

The post US utility expands private LTE network appeared first on Mobile World Live.

Zayo partnered with Nvidia to build out critical network capacity aimed at supporting AI factories, as the digital infrastructure provider looks to close a widening gap between compute demand and available network bandwidth.

Through the collaboration, Zayo is constructing more than 8,000 miles of new long-haul fibre across fast-growing AI corridors, alongside a significant expansion of capacity across its existing infrastructure.

The work spans 24 total routes, comprising six new long-haul routes, 10 overbuilds of existing network across high-demand markets, and eight existing fibre routes.

Nvidia is a Zayo customer and will have significant access to capacity across all the new and overbuilt routes.

Zayo CEO Steve Smith stated AI is reshaping where and how network infrastructure needs to be built across the US, adding the company has invested heavily in modelling where AI-driven demand will emerge so it can expand infrastructure ahead of that demand.

He said Zayo is combining decades of experience building and operating large-scale network infrastructure with Nvidia’s accelerated computing capabilities to support AI connectivity across North America.

Zayo said the expansion is designed to benefit the broader AI ecosystem beyond the largest hyperscalers, including neoclouds, frontier model developers, and enterprises across sectors such as healthcare, finance, and manufacturing.

Over the past 18 months, Zayo has expanded its AI-focused network buildout to more than 15,000 route miles across North America.

Its recently completed acquisition of Crown Castle’s fibre and small cell assets added 90,000 metro route miles and 40,000 on-net enterprise locations, strengthening the company’s metro density for AI inference workloads.

The post Zayo teams with Nvidia to expand AI network capacity appeared first on Mobile World Live.

SpaceX outlined plans to build a massive semiconductor fabrication facility in the US state of Texas, with an initial investment of more than $16.8 billion designed to support growing demand for AI chips across its satellite and space operations.

Texas governor Greg Abbott announced the first phase of the Terafab project, which is expected to create 3,000 jobs and has been backed by a $30 million Texas Enterprise Fund grant. The project also qualifies for incentives through the state’s Jobs, Energy, Technology, and Innovation programme.

The planned 100-million-square-foot site in Grimes County, Texas, will combine logic, memory and advanced packaging under one roof, enabling SpaceX to consolidate parts of the semiconductor production chain as it seeks to avoid constraints in global chip supply.

SpaceX CEO Elon Musk stated the facility will help the company produce AI chips at scale for use on Earth and in space. The project is expected to support future hardware requirements across Starlink, Starship and broader space infrastructure.

Abbott stated Texas was selected because of its business climate, workforce, infrastructure and logistics advantages. Local officials said the development could provide a long-term economic boost for Grimes County and Southeast Texas, with roles spanning engineering, technical operations and plant management.

The investment deepens SpaceX’s presence in Texas, where it already operates its Starbase headquarters, a site in Cameron County and a Starlink manufacturing facility in Bastrop.

SpaceX is pitching Terafab as part of a wider push to secure in-house chip capacity for next-generation aerospace, connectivity and AI workloads, as demand for customised compute grows across the company’s satellite broadband and launch businesses.

Intel also signed up for the project by working with SpaceX, Tesla and xAI to help refactor silicon fab technology at Terafab and support an ambition to produce 1-terrawatt of compute annually for AI and robotics workloads.

The post SpaceX plots $17B AI fab facility in Texas appeared first on Mobile World Live.

AT&T picked vendor partner Ericsson to supply 600MHz dual-band radios as it works to put newly acquired spectrum from EchoStar to use across its wireless network.

The operator stated the move is part of its ongoing effort to boost capacity, coverage, and performance nationwide, with a particular focus on improving reliability in harder-to-reach places, including indoors and in rural areas.

AT&T explained the 600MHz spectrum is especially valuable because it can travel farther and penetrate buildings more effectively than some higher-frequency spectrum, helping customers stay connected in more locations, whether at home, at work, or on the move.

It is more than 60% of the way through a broader wireless network modernisation effort. In areas already upgraded with Ericsson infrastructure, AT&T said customers are seeing up to a twofold improvement in average speeds, a 10% reduction in dropped and blocked calls, and up to a 70% reduction in instances of slower data speeds. Standardising equipment at the top of cell towers has also cut uplink interference by as much as 80%, the operator said.

Ericsson was selected through an open, competitive process to supply the 600MHz dual-band radios, which will enable the operator to deploy uplink-enhancing 8RX technology across low-band spectrum for the first time.

AT&T noted the work is designed to strengthen the network for AI-driven applications, including content creation, sharing, and real-time interaction, building on the companies’ previously announced collaboration to accelerate open RAN deployment.

“We’re more than halfway through our wireless network modernisation, and we’re executing to plan,” said Yigal Elbaz, SVP and Network CTO at AT&T. “By combining our newly acquired 600MHz spectrum with Ericsson’s advanced radios, we’re improving coverage and performance for customers today while building a more modern, flexible network for the future.”

Dyon Agnew, SVP and head of Customer Unit AT&T at Ericsson Americas, said the deployment demonstrates the impact open architectures can have on performance at scale, adding that the two companies are working to deliver faster speeds and more consistent coverage across the US.

AT&T said it will share further details on timing and commercial availability of the 600MHz deployment at a later date.

It received final approval for the $23 billion EchoStar deal in July 2026. AT&T started deploying the 3.45GHz spectrum it bought from EchoStar last year through a short-term spectrum manager lease.

The post AT&T taps Ericsson for 600MHz spectrum rollout appeared first on Mobile World Live.

Google CEO Sundar Pichai (pictured) revealed a major shake-up within its AI division, with DeepMind co-founder and CEO Demis Hassabis stepping back from day-to-day operational management to focus on overarching strategy, AGI and scientific breakthroughs.

Under the new leadership structure, Koray Kavukcuoglu, former CTO of Google DeepMind and chief AI architect, steps up as SVP of Google DeepMind, reporting directly to Pichai.

Kavukcuoglu, who has been with DeepMind for 13 years and previously led foundational projects like WaveNet and Deep Q-Networks, will now directly oversee Gemini model development, frontier AI research and developer ecosystems.

Hassabis transitions into a newly created strategic role as chair of Google DeepMind and chief scientist of Alphabet.

In a note sent to employees, Pichai stated the shift allows Hassabis to dedicate his attention to shaping the long-term future of AGI and scientific applications, alongside his ongoing leadership at drug-discovery subsidiary Isomorphic Labs.

Hassabis expressed confidence in the transition, emphasising  he plans to focus on broad strategic initiatives and high-impact applications of AI in healthcare, particularly accelerating efforts to treat major diseases.

The executive shift comes at a period of rapid product momentum for the internet giant’s AI stack.

According to internal metrics shared by Pichai, the flagship Gemini consumer application has surged past 950 million monthly active users, while the open-weights Gemma model family has surpassed 900 million downloads.

The reorganisation also marks the end of an era for two foundational figures in modern computing infrastructure.

Jeff Dean, Google’s chief scientist and long-time head of Google AI, is departing after 27 years alongside Google senior fellow Sanjay Ghemawat.

The pair are launching independent public benefit company aimed at applying machine learning to scientific and engineering discoveries. Google will serve as a founding investor and cloud infrastructure partner for the new entity, maintaining an active collaborative research framework on advanced systems.

“Jeff Dean is at a moment where he wants to try something new, and we’re excited to support him in that,” Pichai stated.

As competition in model capabilities intensifies across the industry, Google is doubling down on its full-stack approach which spans custom silicon, cloud infrastructure, and consumer applications to seek an edge in the race toward advanced AI.

The post Google DeepMind CEO steps down in exec reshuffle appeared first on Mobile World Live.

SpaceX CEO Elon Musk (pictured) highlighted an ambitious growth plan for the company’s Starlink broadband service, forecasting a hundredfold increase in total delivered bandwidth driven by its next-generation V3 satellite constellation.

Speaking on SpaceX’s first earnings call as a public company, Musk and president Gwynne Shotwell detailed a long-term strategy to capture a majority of global internet traffic in accessible markets.

To support the push, the operator recently applied to the Federal Communications Commission seeking approval to deploy up to 100,000 low-Earth-orbit satellites for its V3 satellite network, supplementing the approximately 9,600 broadband V2 birds currently in orbit.

SpaceX’s Starlink broadband customers doubled to 12 million year-over-year despite price increases for the service. Starlink ended Q2 with service availability in 167 markets “with some very exciting announcements coming soon,” he said.

Those numbers are the tip of the iceberg, according to Musk.  

“This is a big deal,” he said. “It’s kind of hard for people to wrap their minds around this, but it’s not out of the question that at some point, Starlink will deliver a majority of the world’s internet, at least in countries where we’re allowed to operate.”

Musk stated individual Starlink V3 satellites deliver roughly 10 times the performance of V2 models. While V2 hardware typically maxes out between 100Mb/s and 450Mb/s depending on network congestion, the V3 constellation deployment is designed to unlock gigabit-class downlinks for standard end users.

Musk stated a V3 Starlink bird will launch on Starship flight 14, but this time they’ll enter an “operational orbit.” Last month’s Starship flight sent up a batch of V3 satellites as part of a suborbital test.

“The Starlink V3 satellite is about an order of magnitude more capable than the Starlink V2 satellite, which was already a very capable satellite,” he explained. “And we expect to launch about an order of magnitude more Starlink V3 satellites.

Combined with plans to launch 10 times as many V3 birds, the upgraded constellation is projected to increase overall bandwidth capacity by two orders of magnitude. He emphasised that even if revenue per bit decreased by 90%, the expansion will still drive a tenfold increase in Starlink revenue.

“It’s certainly a bold claim and one that contravenes the company’s prior statements that Starlink was most appropriate for rural users,” Quilty Space founder and CEO Chris Quilty told Mobile World Live. “Starlink V3 will provide a 10x boost, reaching 8Pb/s of downlink capacity, but aggregate network capacity is a misleading metric because capacity is geographically distributed while demand is geographically concentrated.”

AI compute capacity
SpaceX ended Q2 with 1.4 gigawatts of compute capacity, up from 1 gigawatt in the prior quarter and 400 megawatts a year ago.

“We expect to end this year with over 2 gigawatts of compute,” Musk stated.

Asked about the outlook for compute capacity, Musk revised his earlier target upward on the call. Rather than the roughly 10 gigawatts previously floated for the end of 2027, he said SpaceX is now aiming for approximately 15 gigawatts of power and cooling capacity online by the end of next year, with a stretch goal of 20 gigawatts.

“We’re actually aiming to far exceed that gigawatt number in terms of power online, power cooling, and electrical equipment,” he explained.

“I don’t think we’re going to achieve 20 gigawatts, but we want to have a series of projects that cumulatively come to 20 gigawatts by the end of next year,” he added. “Our goal is to have far more power, cooling, and electrical equipment than we have GPUs. That’s a logical thing to do given the relative expense of GPUs.”

He said SpaceX is taking the expertise it uses for rockets and satellites and applying it to scaling terrestrial data centres.

“SpaceX as a company is incredibly good at hardware,” Musk said.

Going forward on data centres, Musk said SpaceX has decided “to build exclusively on Nvidia, because we think the Vera Rubin is the best architecture.”

The post Musk targets global broadband domination with Starlink appeared first on Mobile World Live.

Anthropic confirmed it is building an in-house silicon team to design custom chips for its Claude AI model, a move aimed at improving performance and efficiency as demand for its services continues to rise.

In a statement to Mobile World Live (MWL), Anthropic said it “is hiring engineers with experience across the stack”.

The company stated this move will let it co-design hardware and models together, allowing Claude to run faster and more efficiently at the scale its customers need.

Anthropic told MWL the effort is not a shift away from its existing hardware partners. The company described the silicon team as the latest step in its multi-chip approach

It is building a diversified hardware stack with chips from Amazon Web Services (AWS), Google, Nvidia, and AMD, which will “remain central to how we scale”.

Reuters had previously reported Anthropic was exploring building its own chips while The Information stated the company had held talks with Samsung Electronics about possibly manufacturing them.

Business Insider cited a newly posted job listing which describes what Anthropic is looking for, which includes a “custom silicon team” with engineers spanning multiple areas of chip design and verification.

The listing, which advertises a salary range of $320,000 to $485,000, calls for candidates who can point to direct personal contribution in finalising and shipping semiconductor designs.

Building AI chips in-house is expensive and technically demanding, but the benefits include improved performance and less dependence on Nvidia.

Rival moves
Anthropic joins a growing list of AI companies pursuing custom silicon, following moves by OpenAI, Meta, Google, and Amazon.

OpenAI unveiled its first custom-built inference ASIC, Jalapeno, in June, developed in partnership with Broadcom.

In April, Meta Platforms expanded its partnership with Broadcom to jointly develop multiple generations of custom MTIA chips, building on four new generations it had unveiled in March.

Google is also developing a server chip, called Frozen v2, to improve efficiency for its Gemini AI model and separate from its existing TPUs,

During Alphabet’s Q1 earnings call in April, CEO Sundar Pichai stated the company will begin making its tensor processing units available to a select group of customers for their own data centres, a move seen as a potential challenge to Nvidia’s dominance in AI chips.

Also in April, Meta struck a deal with Amazon Web Services to use its Graviton processors for agentic AI workloads.

Amazon also has an agreement in place with Anthropic to use its Trainium chips to train and run its AI models

The post Anthropic plans in-house chip team to power Claude appeared first on Mobile World Live.

SpaceX president Gwynne Shotwell claimed Starlink Mobile is positioned to compete directly with AT&T, Verizon and T-Mobile US, pointing to the trio’s roughly $600 billion in combined annual revenue as a measure of the opportunity.

Speaking on SpaceX’s Q2 earnings call, Shotwell said the service’s capability could improve roughly 100-fold once the company deploys next-generation birds and spectrum acquired from EchoStar, with commercial service targeted to begin by the end of next year.

Shotwell said the current direct-to-cell system operates using around 5MHz of bandwidth through local telco partners, but SpaceX now has access to 65MHz of bandwidth via the EchoStar spectrum deal, alongside plans to deploy roughly 10 times as many satellites.

“So, simplifying, you could look at Starlink Mobile leveraging the next-generation satellite and the EchoStar spectrum as being 100 times better, 10 times 10 better than what we’ve got right now,” she said, adding the upgraded service already supports voice and video calls over apps such as Signal and WhatsApp.

Shotwell described a terrestrial deployment approach built around Starlink’s existing broadband hardware rather than a conventional cell tower build-out.

“You could put a base station, a cellular base station on the gear that holds a Starlink broadband dish,” she said. “You can have kind of these little femtocells around the country and you deploy that as you need it. You don’t have to necessarily spend many billions of dollars in low band spectrum all upfront before you deploy your system.”

She was reluctant to give a specific capex figure for the terrestrial build-out.

“We have a lot of really kind of great and new ideas for how we’re going to do that,” she said. “And I think it will be quite capex efficient.”

CEO Elon Musk elaborated on the same point, saying it was worth emphasising given its importance to the economics of the rollout.

Rather than deploying expensive, hard-to-site large cellular base stations, he said SpaceX expects to deploy a large number of small stations which are effectively Starlink dishes also equipped to provide connectivity in mobile spectrum bands.

“The Starlink antennas are located on the roofs of houses and businesses,” he said. “So, they’ve got a sort of a clear view for providing connectivity directly to cell phones on the ground.”

Musk expects the resulting service to offer “better and higher bandwidth than what is currently available from cellular providers”.

Shotwell stated SpaceX expects strong customer uptake once the improved service launches.

“I anticipate us to be able to acquire quite a few of their customers, because I think our service will be better,” she said, citing Starlink Mobile’s ability to eliminate dead zones and maintain service during natural disasters, given the relative stability of the satellite environment compared with terrestrial infrastructure on the ground.

Q2 numbers
SpaceX reported Q2 revenue of $7.8 billion, up 92 per cent year-on-year from $4.1 billion, in the company’s first results release since going public in June.

Net loss narrowed to $541 million, an improvement of $467 million from the $1 billion loss posted a year earlier.

Capex surged to $18.4 billion for the quarter, up from $2.8 billion a year earlier, driven overwhelmingly by the AI segment where it reached $15.8 billion. Space segment capex came in at $1.2 billion and Connectivity capex at $1.4 billion.

The AI division, spanning Grok, X, cloud infrastructure and advertising, posted revenue of $2.6 billion, up 247%, driven by new Cloud Services Agreements and Grok/X subscription growth.

The business generates compute-contract revenue from customers including Anthropic, Google and Reflection AI, though some recurring revenue has yet to be recognised.

The segment narrowed its operating loss 49% quarter-on-quarter to $1.3 billion, though SpaceX cautioned the unit needs sustained investment to turn profitable. It signed $14.1 billion in contracted Cloud Services Agreements and announced a $60 billion deal to acquire coding tool Cursor.

Connectivity, the Starlink-led segment, generated $4.3 billion in revenue, up 66%, with operating income up 79% to $1.7 billion. Starlink subscribers doubled to 12 million.

The Space segment posted revenue of $962 million, up 29% with an operating loss of $542 million as the company invested heavily in Starship.

SpaceX ended the quarter with $100 billion in cash and marketable securities, and a $47.5 billion backlog.

The post SpaceX claims Starlink Mobile will rival AT&T, Verizon, T-Mobile US appeared first on Mobile World Live.

Chipmaker Intel joined the OCUDU Ecosystem Foundation as a premier member, bolstering open-source RAN software development for 5G and future 6G networks.

Hosted by the Linux Foundation, the OCUDU Ecosystem Foundation operates as a vendor-neutral collaboration hub.

Its core mission is to establish reference architectures, validation programmes and continuous integration and testing pipelines designed to reduce integration complexity and streamline production rollouts for operators.

Intel’s entry brings technical weight to the initiative, particularly across cloud-native infrastructure, silicon platforms and telecommunications architecture.

Intel plans to contribute technical capabilities to help shape open, standards-based solutions and advance AI-native reference implementations capable of running on software-defined infrastructure.

Arpit Joshipura, GM of networking, edge, and IoT at the Linux Foundation, stated Intel’s leadership in cloud-native systems and telecommunications will bolster ongoing efforts to build an open, interoperable ecosystem, ultimately allowing operators to deploy AI-native open-source RAN solutions faster and with greater confidence as the market shifts from innovation to production.

The foundation’s technical work focuses heavily on the central processing components of open wireless networks: the central unit (CU) and distributed unit (DU). By providing a shared, software-defined reference platform built on open-source code, the initiative aims to address multi-vendor interoperability bottlenecks and eliminate the risk of proprietary vendor lock-in.

Intel joins a growing coalition of communication service providers, infrastructure vendors, system integrators and research institutions backing the open-source initiative following its launch earlier this year.

The foundation continues to invite operators and technology partners to participate in its technical working groups and contribute directly to open reference code, accelerating the transition toward scalable, production-ready open RAN deployments.

The post OCUDU group signs-up Intel to boost open RAN push appeared first on Mobile World Live.

Anthropic appointed Mariano-Florentino Cuellar as its first chief global affairs officer, as the Claude AI model developer faces escalating political and regulatory friction.

In the newly-created role, Cuellar (pictured) will report directly to Anthropic president Daniela Amodei.

His primary focus involves managing relationships with US authorities and international government bodies as the business expands internationally.

Cuellar is stepping down from Anthropic’s independent four-person Long-Term Benefit Trust, a body which ensures the company aligns its commercial goals with its public-interest mission.

The trust is now searching for a successor.

Anthropic described Cuellar as a pragmatist and consensus builder suited to navigating government opposition. His credentials include working as a special assistant to former President Barack Obama and a justice with the Supreme court of California.

He also directed the Carnegie Endowment for International Peace think tank while co-chairing a task force on US national security and nuclear proliferation, a domain industry executives often cite as a blueprint for managing advanced AI risk.

Cuellar also jointly led studies which shaped landmark Californian state law SB 53, which introduced whistleblower protections and safety reporting mandates for high-revenue AI developers.

The executive said politicians everywhere are recognising AI governance is at a pivotal inflection point, noting current decisions would determine if technology can be harnessed safely or increase global inequality.

Anthropic is working to mend relations with the US government following several clashes including a military dispute which resulted in the US Department of War blacklisting its technology, a designation the company is challenging in court.

Officials also issued export control directives temporarily banning international sales of its high-tier Mythos 5 and Fable 5 models over national security concerns.

The post Anthropic recruits first global affairs chief appeared first on Mobile World Live.

Global smartphone revenue rose 7% year-on-year in Q2 to an historic $109 billion, the highest second-quarter total on record even as shipments declined, Counterpoint Research revealed.

The disconnect between rising revenue and falling volumes was driven by average selling prices (ASPs), which rose 17% to a record high of $400, fuelled by sustained demand for premium devices and memory-driven price hikes across Android portfolios.

Apple was the standout performer, with revenue up 22% to the brand’s highest-ever Q2 total.

The growth was supported by a 13% rise in shipments alongside an 8% increase in ASP, pushing Apple’s revenue share to a record 49%.

Counterpoint Research director Tarun Pathak said Apple’s performance was underpinned by sustained demand for the iPhone 17 line-up, particularly the base 17 and 17 Pro Max, which kept the sales mix skewed toward premium smartphones.

Unlike rivals which pushed through steep price increases, Apple largely held pricing steady, a move Pathak said reflects the company’s ability to absorb rising bill-of-materials costs and stay comparatively insulated from the industry’s memory chip crunch.

He cautioned Apple would likely raise prices in coming quarters, but believes the pricing discipline so far has strengthened its competitive position, enabling growth in both value and volume.

China, Europe and emerging markets were strongholds, as Apple’s stable pricing improved its relative value proposition against widespread Android price hikes.

Samsung ranked second with a 16% revenue share, with both revenue and shipments up 9% helped by steady Galaxy A-series demand and continued momentum from the Galaxy S26 series, along with strong growth in North America and the Middle East and Africa.

Xiaomi posted the steepest shipment decline among the top five brands, down 26%, dragging revenue down 17% despite a 13% rise in ASP as the company pushed toward higher-value devices.

Oppo and vivo also saw revenue fall, down 10% and 11%, respectively.

Senior analyst Shilpi Jain said the market has entered a phase where value, not shipment volume, is the primary growth driver, with OEMs increasingly passing higher component costs on to consumers.

With memory shortages showing no near-term signs of easing, the research company expects further price increases and steeper shipment declines across the industry in the second half of 2026.

The post Smartphone revenue hits $109B as Apple claims record share appeared first on Mobile World Live.