The oversubscribed, sustainability-linked financing will help fund its EMEA data centre footprint
Global data centre operator, EdgeConneX, has secured an additional $1.9 billion in sustainability-linked financing to support its EMEA digital infrastructure expansion. The latest deal, which was marked by significant oversubscription from existing and new lenders, adds to EdgeConneX’s $2.9 billion in sustainability-linked financing, initially issued in October 2022.
With a growing European footprint of more than 20 facilities, EdgeConneX continues to expand its presence in the region, driven by strong global customer demand for cloud and AI infrastructure.
The Swedish-based infrastructure investment company EQT-owned data centre operator has now raised more than $7 billion in total green investments over the past two years.
The financing consolidates EdgeConneX portfolio across EMEA, transitioning it from stand-alone project-based funding to a more robust structured corporate debt package. The company said this latest transaction provides it with “enhanced financial flexibility” and a greater capacity to support projected growth throughout the EMEA region.
Part of the latest deal includes sustainability-linked margin adjustments. This feature directly ties the interest rate to EdgeConneX achieving predetermined sustainability objectives. While PUE measures sometimes do not work that well for operational data and multitenant data centres, to its credit, EdgeConneX is also looking to eliminate its waste and water footprint as it moves to powering its entire data centre platform with renewable energy sources by 2030.
Tripled capacity
Since EQT’s acquisition in 2020, EdgeConneX has more than tripled its capacity and expanded into Asia, Latin America and new European markets. Today, the company has a global footprint of 80 data centres in operation or development in more than 50 markets across North America, Europe, APAC and South America.
“The reception of this transaction reconfirms our status as an industry leading data centre developer and operator and reinforces our belief that responsible growth and operational excellence go hand in hand,” said EdgeConneX CFO Joe Harar.
“It showcases our ability to replicate this success on a global scale. This is a significant milestone in our journey towards leading the industry in sustainable practices, and we are excited about the future possibilities this opens up for EdgeConneX and our stakeholders,” he added.
“With this latest sustainability-linked financing deal, EdgeConneX is rapidly ascending as one of the largest financing issuers in Europe, underscoring our commitment to sustainability and innovation in the data centre sector,” said EdgeConneX SVP corpoate finance Eelco Holst. “Our team’s exceptional achievement was accomplished through dedicated efforts in cultivating a broad range of lending relationships, as well as close cooperation with the coordinators, advisors and EQT Infrastructure.”
Partner content: Encryption of digital comms protects individuals’ privacy, but obscures payloads and endpoints so law enforcers have shifted their investigative focus
The encryption of nearly all digital communications today advances individual privacy protections, but it also obscures payloads and endpoints from lawful interception. Accordingly, law enforcement agencies (LEAs) have shifted their investigative focus from message payloads to traffic flows and distributed data sources. This class of evidence increasingly depends on locating, assembling, and drawing insights from digital breadcrumbs, rather than from directly intercepted communications. The breadcrumbs themselves typically require a degree of forensics to locate and identify, followed by additional analysis to reveal their significance and investigative value.
SS8 lawful intelligence tools provide superior traffic analysis based on deep packet inspection (DPI) to identify and aggregate digital markers that advance investigations. These measures illuminate application- and sub-application layer information about who is communicating, when, and across what platform, then enrich that session-specific metadata by associating it with other information. These connected breadcrumbs can reveal insights that help identify individuals, build patterns of life, and expand the network of subjects of interest in an investigation.
Building a picture from breadcrumbs
One way that lawful intelligence commonly advances investigations is by building a network of relationships outward from known provisioning targets being investigated under a warrant. For example, a provisioning target may exchange messages with any number of other individuals, some of whom may be known to investigators as subjects of interest. In most cases, however, at least some of these other parties are unknown, and learning their identities can be valuable.
SS8’s Enhanced Protocol Extraction Engine (EPXE) analyzes communication application protocols and ports, including encrypted flows, to reveal digital breadcrumbs such as an IP address or phone number. Data from a communications service provider (CSP) or existing evidence can then be used to reveal who the B-party is – if a warrant can be obtained. Often, however, those measures are not sufficient, such as if a subject of interest is using a burner phone, and the B-party remains hidden. In such cases, investigators must use other measures to build a case to provision the B-party as a lawful interception target. The SS8 platform provides mechanisms to help compile information about such individuals using publicly available data.
Data fusion solutions
Data fusion solutions like SS8’s MetaHub can automatically investigate connections between known breadcrumbs and additional details, largely using open source intelligence (OSINT) feeds. As a simple example, the B-party’s phone number may show connections to social media accounts or email addresses. Each such connection expands the potential avenues to discover more about the subject of interest. Lawful intelligence collects, evaluates, and analyzes this information from sources that range from ad-targeting, gaming, and blog platforms to tax and land-ownership records.
OSINT-driven investigation is adept at revealing associations between a single person’s multiple online identities. Different social media accounts may reflect various aspects of an individual’s personal, political, and professional selves. Discovering an associated Skype account may reveal an alternate phone number, location, and profile picture. Vehicle records may provide access to toll records and automated number plate recognition (ANPR) systems.
Collecting and collating these breadcrumbs to identify an unknown individual may be referred to as recursive identity lookup. It can help provide a richer understanding of those subjects of interest, including potential connections to the larger investigation, and justify provisioning additional interception targets. EXPE can further refine a target’s pattern of life by identifying which apps they use most and whether they communicate using voice, video, or chat, and MetaHub can be configured to automatically scan OSINT feeds and other available sources to discover new breadcrumbs. In this way, recursive identity lookup supports a rich investigative narrative populated by recognized actors.
Assembling and applying target profiles
Combining multiple internet identities into a single, enriched profile of a subject of interest, rather than analyzing disconnected threads of a digital presence, is critical to investigating an individual. It gives investigators a more complete understanding of the subject and his or her relationships with other people, places, and events. To formalize that model and extend its benefits throughout the lawful intelligence practice, SS8’s Intellego XT includes the iDossier module, which collects important details about a person or organization of interest and flexibly includes those that may or may not be under investigation or provisioned as interception targets.
Details contained within profiles may come from recursive identity lookup, case or investigative records, analyst entries, or any other source. A comprehensive iDossier profile can include a wealth of diverse data such as pictures, online identities, residence and work addresses, physical attributes, voice samples, and other details. They can also be linked to external data sources such as criminal records, passports, and driver licenses. This information can be continually updated to facilitate building rich patterns of life in real-time, with connections among associated individuals and organizations.
Within the broader context of lawful intelligence, the SS8 platform can make iDossier profiles available to analysts with automatic name-number lookup to identify and give context to parties of interest that are communicating with each other. Integrated with the rest of the SS8 platform for capabilities such as analytics, data fusion, and location intelligence, such insights into subject identities help move investigations toward successful outcomes.
About David Anstiss
David Anstiss is Director of Solution Engineering at SS8 Networks. He has been with SS8 since 2015 and has significant experience in critical network architecture technology and advanced data analytics. He currently works as part of the Technical CTO Group under the leadership of Dr. Cemal Dikmen and is responsible for leading engagement with both intelligence agencies and Communication Service Providers (CSPs) around the world. He has been instrumental in helping them transition to 5G, defining system requirements to meet regulatory compliance. As a member of ETSI, he represents SS8 to ensure the adoption of cloud-native infrastructure is met with industry best practices and to guarantee that compliance of lawful interception is maintained. Learn more about David here on his LinkedIn profile.
About Rory Quann
Rory Quann is Head of International Sales at SS8 Networks and brings with him over 10 years of experience in the Lawful Interception and Data Analysis industry. Prior to joining SS8 in 2013, Rory worked for BAE System Applied Intelligence where he was focused on large scale Government deployments of Intelligence Solutions. Rory has held multiple positions in the Lawful Intelligence space ranging from Deployment Engineer, System Consultant, and Sales Engineer focusing on Country-wide Passive deployments. Rory is a Certified Microsoft MCSA Engineer and EMC Certified deployment Engineer. You can learn more about Rory on his LinkedIn profile by clicking here.
Partner content: Telecoms must navigate the complex interplay between the network, automation and AI as all three evolve
In the heart of the digital revolution, the telecoms industry finds itself navigating the complex interplay between traditional automation techniques and the emerging potential of Artificial Intelligence (AI). This landscape presents a unique set of challenges and opportunities for telecom service providers, sparking a critical examination of conventional approaches and the pathways to future advancements.
The Critical Role of Automation
The telecom sector, propelled by the next generation of fixed and mobile services, faces an undeniable complexity. This complexity necessitates a shift towards automation and orchestration, especially in environments where manual processes cannot keep pace with the demands of dynamically orchestrated, virtualised 5G networks. Automation emerges not just as a beneficial tool but as an imperative for maintaining efficient, cost-effective service lifecycle management amidst this evolving backdrop.
Reassessing the Relationship Between AI and Automation
An important consideration in the current discourse is the distinct role of AI in relation to automation. While acknowledging AI’s capacity to transform future operations, it’s crucial to understand that automation’s efficacy isn’t contingent on the immediate adoption of AI technologies. This realisation encourages service providers to tap into existing automation capabilities, fostering innovation and creating immediate value without necessarily depending on AI’s progress. It posits that AI, although promising, serves as an optional enhancement for many telecom processes rather than a foundational necessity.
Navigating Automation Without Immediate AI Integration
Exploring the practical application of automation, independent of AI, reveals how today’s technology can address complex processes, enhance efficiency, and reduce operational costs. This approach covers a spectrum of telecom operations, from service provisioning to enhancing customer experiences, showcasing the tangible benefits of automation leveraged through current technological advancements.
Future Prospects: AI as an Operational Enhancer
Looking ahead, AI is poised to play a significant role in refining telecom operations, capable of optimising complex decision-making processes. However, equating AI as the sole solution for present challenges overlooks the immediate benefits and practicalities of current automation technologies. By adopting a balanced strategy that prioritises automation now, telecom providers can establish a solid foundation for seamlessly integrating AI technologies in the future, ensuring it complements an already efficient operational framework.
Strategic Implementation: A Forward-Looking Approach
The strategy for embracing automation today, without the immediate inclusion of AI, emphasises the significance of practical, actionable measures. It suggests a focused direction for telecom operations, highlighting the importance of adopting automation with a vision that anticipates the eventual integration of AI. This approach not only addresses immediate operational needs but also prepares the groundwork for future technological advancements, promising a strategic advantage in the evolving telecom landscape.
Conclusion: The Imperative for Immediate Action
The discussion culminates in a strong endorsement for the proactive adoption of automation technologies in the telecom industry. This approach presents automation as an essential, readily available tool capable of delivering substantial benefits today, thereby setting the stage for the future role of AI. Such a pragmatic strategy ensures immediate improvements in efficiency and productivity, offering a rapid return on investment while preparing operations for future innovations.
In essence, the telecom industry’s journey through digital transformation is marked by the strategic embrace of automation technologies. This path does not merely respond to today’s challenges but also anticipates the integration of AI, ensuring sustainable, long-term innovation and success. For telecom leaders, the message is clear: the time to act is now. By prioritising automation, the industry can maintain its competitive edge and navigate the complexities of the digital age with confidence and strategic foresight.
DOWNLOAD THE FULL WHITE PAPER – AI-based automation: Wait or act now?
In essence, “AI-based Automation: Wait or Act Now?” is more than a white paper; it’s a strategic manual for service providers navigating the complexities of the modern telecom landscape. It argues convincingly for a balanced, forward-looking approach to automation and AI, offering a blueprint for immediate action and long-term success.
For those ready to chart their course in this dynamic environment, the insights and recommendations within this white paper are invaluable. It’s a clarion call to act now, harness the power of automation, and prepare for the AI-enhanced future. Download the white paper today and take the first step toward transforming your telecom operations with We Are CORTEX.
The draft strives to be a careful balancing act, intended to provide incentives for rivals to build competing fibre infrastructure
Germany’s telecoms regulator, the Bundesnetzagentur (BNetzA) has published its draft approval of the prices Telekom can charge competitors for the use of its fibre ducts and poles. Telekom is Deutsche Telekom’s opco in Germany and the largest network operator.
Klaus Müller, President of Bnetza, commented, “In mid-2022 we imposed an obligation on Telekom to open up its ducts to competitors wanting to deploy fibre. We have now assessed Telekom’s proposed prices for access to its ducts. Our aim has been to accommodate Telekom’s legitimate interests while enabling competitors to realise their business models.
“This balance of interests creates incentives to invest in swift fibre rollout across the country for the benefit of all consumers.”
Lower prices
The BNetzA assessed a range of prices proposed by Telekom and is considering lowering many of them, and in some cases making them much lower.
This is the first time a price assessment has taken into account Telekom’s dominant position to ensure that there is still an incentive for competitors to roll out fibre networks.
The prices calculated by Telekom factored in its potential loss of customers and concomitant loss of revenue to competitors – known as lost contribution margins. The regulator followed the same approach but also took into account the costs incurred by Telekom.
It took the view that marking up costs to cover the potential loss of customers is only applicable to Telekom’s investment in fibre infrastructure. For example, connections between street cabinets and customers, where Telekom has only just begun to deploy fibre.
The regulator says is not yet possible to predict exactly how the take-up of duct access will pan out and hence the initial price approval will only last two years: Europe’s largest economy has one of the lowest fibre penetration rates on the continent.
The German regulator and all network builders would do well to examine what has happened in the Spain, which is leagues ahead of all the other large countries in Europe regarding fibre penetration and overbuild. You can read Nae’s illuminating analysis here.
Package of measures
BNetzA’s decision on pricing is only one of a series of decisions relating to duct access.
The regulator has obliged Telekom to submit its contractual terms and conditions for the use of its civil engineering assets (its “reference offer”) for assessment.
In addition, Telekom must provide data on available duct capacity on a regular basis for publication in the Infrastructure Atlas. Telekom has so far met these obligations and the regulator says it will announce its decision on Telekom’s reference offer “in due course”.
Next steps
Comments on the draft approval decision can be submitted until 10 May 2024. The final decision will follow a consultation process that involves the European Commission and the regulatory authorities of the other EU countries.
The move follows KPN’s acquisition of the company’s Dutch operations
Belgian broadband and MVNO Youfone, which buys wholesale services from Proximus, has announced it is rebranding as Yoin after KPN finally received the “all-clear” from the Netherlands Authority for Consumers and Markets (ACM) to taken over Youfone’s Dutch operations. The name change will be official on 15 April.
The rebranding further suggests KPN intends to continue using the Youfone brand in the Netherlands. At the close of that deal KPN intimated that was the plan anyway: “The Youfone organization will continue to operate independently within KPN to ensure that current and future Youfone customers are served in the way they are familiar with.”
Youfone began its life in 2008 in Rotterdam. In 2021, Proximus signed a wholesale agreement with Youfone, which at the time claimed to be the largest independent operators in the Netherlands with over 400,000 customers, for mobile services. That number has risen to 580,000 today but this figure is just the Dutch subscribers, not the Belgian side.
In parallel, the two also signed a wholesale agreement for fixed internet access, allowing the Dutch operator to add fixed services to its mobile offers, which the operator began in April that year.
No changes to its offers
In a statement, Youfone/Yoin said: “You can also expect the many benefits you received at Youfone at Yoin. You can still use the friend service. You and your friend will both receive a €10 discount on your next invoice. The MyYoufone app was also given a new look and is now the MyYoin app. The app gives you a clear overview of your usage and invoices and you can set a calling and data limit.”
The operator is already one of the cheapest in the Belgian market and looks like it will continue on that path, suggesting that with its current 50% off for three months offers, the cheapest subscription at Youfone starts at €5 per month.
It is unclear whether the Youfone cows (above) will reappear as part of the new branding.
Benelux shakedown
KPN’s acquisition of Youfone and the subsequent rebrand of the independent Belgian operations to Yoin marks the second big rebranding in the region. Last September, T-Mobile Netherlands and Tele2 Mobiel adopted a new brand as independent Dutch brand Odido. The telco’s brands Simpel, Ben and Tele2 Thuis also fall under Odido and continue to exist. Odido has more than 2,000 employees and 119 stores in the Netherlands.
That rebranding followed the telco’s acquisition, in September 2021, for more than €5 billion by private investors Apax and Warburg Pincus. T-Mobile took over all of Ben’s shares in 2002 and later merged with the Dutch branches of Orange and Tele2.
In July last year new MVNO Undo announced it was partnering Orange Belgium to offer its climate action focused brand in the market. Undo’s mobile packages start at €10/month and include 20 GB/500 minutes/100 SMS, plus “2kg of CO2 to compensate” as part of carbon offsetting.
Once complete, the DC operator will have 64MW of critical IT capacity in Switzerland
Global data centre operator Vantage Data Centers announced that it will open a second Zurich campus (ZRH2) this summer as it continues to expand in Europe. The new carrier-neutral campus, which will provide 24MW of critical IT capacity, is located 30 kilometres north of Zurich’s City Center in Glattfelden. It is situated just over 20 kilometres from Vantage’s ZRH1 campus in Winterthur and 15 minutes from Zurich Airport.
With this campus being the company’s 33rd globally, it demonstrates just how in the current AI gold rush, the companies supplying the picks and shovels continue to thrive and secure well-heeled backers.
Last October, Vantage completed a €2.5bn investment partnership with consortium of investors led by MEAG and Infranity, along with funds affiliated with the investment management platform of DigitalBridge, to build six stabilized data centres in strategic locations across Europe. The data centres span more than 1.8 million gross square feet and 177MW of IT capacity.
The new ZRH2 will have a total of 226,000 square feet (21,000 sqm) of space for Vantage’s customers consisting of the usual suspects. ZRH2 will also offer 300W/SF (3.2kW/sqm) average densities.
“Zurich is a rapidly growing data centre market due to its strategic central location and its global financial leadership, and we expect to continue developing in this market based on customer demand,” said Vantage Data Centers president EMEA David Howson.
“Across Europe and the globe, the demand for highly efficient, hyperscale data centres is booming to meet the needs of next generation applications, from powerful AI models to high-performance computing and cloud transformation. Vantage is prepared to meet that demand for current and new customers, and we look forward to opening the doors of our latest data centre this summer,” he added.
More specs
Major fibre-based service providers are operating within a few miles of the data centre and Vantage said the facility will have two Meet-Me-Rooms allowing for “diverse paths and multiple connectivity options.” This includes two points-of-entry (POEs) ensuring maximum path diversity for inbound carriers (minimum two paths per carrier).
Vantage Data Centers has committed to achieving net zero carbon emissions across its global portfolio by 2030 and said ZRH2 will deliver “industry-leading” power usage effectiveness (PUE) and water usage effectiveness (WUE) to support energy-efficient operations.
The operator has installed air source (ASHP) and water source (WSHP) heat pumps within ZRH2. These heat pumps will convert waste heat for use to condition administrative spaces and reduce external energy use. ZRH2 will also provide waste heat to a nearby hotel and seminar centre. Additionally, ZRH2 will utilise rainwater infiltration and a green roof to minimize environmental impacts.
Dr Richard Windsor of Radio Free Mobile slices through the hype, highlighting the importance of causal understanding, which LLMs lack
In his latest blog, Dr Richard Windsor, Founder and of Radio Free Mobile (RFM) is not impressed by Meta and OpenAI’s claims that their next models will be able to reason.
If the claims were true, it would be a huge step towards “super-intelligent machines”. As it is, he reckons the models will simulate reasoning, like all those models that have gone before, which is not the same thing at all.
Causal vs correlation
For the last six years RFM Research has argued that the main limitation of AI systems is that they are based on deep learning and have no causal understanding (perception of cause and effect).
Rather, they are all sophisticated pattern recognition systems that can identify correlation but cannot decide whether a relationship is causal or simply correlated.
This flaw affects all kinds of AI systems, from the simplest neural networks to the biggest large language models (LLMs).
Despite this, many AI actors claimed the current crop of LLMs have the power to reason, yet there is much fanfare about their successors having that capacity.
The Financial Times [subscription needed) reported on the claims for the next generation of AI engines. Dr Windsor thinks Meta’s commentary cited in the article is more realistic than that of OpenAI’s in the said article because it expresses aspiration.
Lies, damn lies and statistics?
On the other hand, he thinks OpenAI seems to imply that it can implement reasoning via a purely statistical-based system and thinks this unlikely on the grounds that, “as the only way I know of achieving reasoning in computer systems is to use the rules-based software that has been around for years.
“The reason why a cheap pocket calculator is better at maths than a multi-billion dollar LLM is because it is programmed with the rules of maths whereas the LLM uses statistics and probability to achieve its results.
“In short, software can reason but it can’t learn whereas deep learning can learn but it can’t reason.”
LLM + software =
If we accept this statement as being true, then, Dr Windsor argues, “it stands to reason that if one were to combine software with an LLM, this would represent an effective and efficient method of implementing reasoning in these systems.”
In case you were wondering, this is referred to as neuro-symbolic AI, Dr Windsor explains – “symbolic is a fancy term for software” and “was a relatively active area of AI research 4 years ago (see here) but it has been pretty quiet since all the focus moved to LLMs”.
So “Unless OpenAI is referring to using software in its LLMs to provide the reasoning, then I am certain that GPT-5, LlaMa 3 and so on will be as incapable of reasoning as all of their predecessors,” the doctor says.
He agrees these systems are “very good at simulating reasoning” but given real empirical reasoning tests, “they fail and fail convincingly”.
Not the end of civilisation
And concludes that “despite the chatter and the hyperbole, I continue to think that we remain as far away from super-intelligent machines and artificial general intelligence as we have ever been”.
This is important for a number of reasons. Perhaps most importantly to the biggest number of people, “hundreds of millions of jobs are safe for the foreseeable future and that the robots are not coming to kill us yet”.
“We are in an AI bubble and at some point, everyone will realise what this new branch of AI can realistically do and what it is just noise,” Dr Windsor says.
The operator panellists shared their candid experiences around how fully integrated telcos can transition to NetCo, ServCo and InfraCo models that work
The discussion took place at our recent Telecom Europe Telco to Techco virtual event, moderated by Omdia practice leader James Crawshaw. Our panellists were Telenor cloud strategy and architecture director Pål Grønsund, Veon Group chief digital operations Lasha Tabidze and Redhat global account manager Brian Klafstad.
Grønsund explained that three models are emerging for service providers: Infraco which is the towers and fibre; NetCo which is network-as-a-service; and ServCo which is connectivity as a service. Telenor split passive infrastructure from active infrastructure but in some markets where it operates, some of the joint ventures include active infrastructure. “It is really important to look at the specific business ambitions and market conditions as well to see what out of those models would be the better fit,” he said.
Tibidze said operators separate to gain efficiencies or create value. He said Veon’s approach is to potentially sell a portion of the TowerCo/NetCo to free up capital to reallocate to the ServCo business, whether that is service delivery, AI or digital services.
Different investor classes
Klafstad highlighted that each model attracts different investor classes. “The ones that invest in your your ServCo will be wanting a much higher margin, because there used to the EBITDA, so they’re looking at 30%+ margins,” he said, adding that InfraCo investors are long-term and can be happy with 5%+ margins. He said telcos must explore efficiencies and automation here and realise that each new telco layer moves closer to customers so can sell more things to them, rather than just space.
For example, Telenor Infrastructure has three focuses now, the sites (including building and power), the fibre and the data centre. “The key motivation is to focus on what is the utilisation of these assets and then that could be attracting new customers, it could be attracting external investors as we have talked about. Another important thing is also how to improve the main processes, dealing with those assets and managing them,” said Grønsund.
Not just cost savings for the split
Veon’s digital operator strategy, according to Tibidze, is not about cost saving at all. It’s not just about selling off towers, it’s more about creating a dedicated organisation that’s looking after that infrastructure within the telco with its own p&l. Then they have the right incentives to run their business segment more efficiently than when it was just a cost centre inside of a larger telco organisation.
“TowerCos when separated are more focused and can move into things like green energy – it’s not about just new clients on your towers,” he said. Therefore changing the telco business model is “not initially done because of the cost saving; it’s initially done because of specialisation and increasing the efficiency.”
“We clearly see that dedicated specialised teams or companies are really doing much better on the passive infrastructure side,” he added. “Of course, you can keep this company within your umbrella and structure. But there is a huge possibility of finding the investor, going into the JV or even selling to specialised companies, all this infrastructure.”
• Where edge services fit into the bigger picture and what the business model might look like including the opportunity for CPU-as-a-service
• Creating sources of funds to grow your ServCo business
• New InfraCo services which are working in the market
• Different investor models emerging for the different telco layers
• Attracting new customers onto your various asset classes
• Infracos own many sites around in the markets where they operate. That’s an opportunity for ServCos to build at the edge but also for others to come in and put their edge cloud into those locations
• Sovereign cloud opportunities and rising requirments in this space from the likes of government and defence
• Veon’s lessons with its Bangladesh operational restructure
• How the ServCo/NetCo/InfraCo model allows each segment to create more value through their focused investment decisions beyond what an integrated telco can do given it must make judgements on everything from building and conduits right though to investing in GenAI
• Lessons from Danish incumbents TDC’s painful structural separation
• Strategies to successfully migrate from legacy IT to open stack software despite consensus believing this is difficult for telcos to pull off
• Building the right teams to make these more-focused organisations work better
• How a more focused business and management has helped Telenor and Veon with emerging edge services and what these may look like
Telco says Google Distributed Cloud (GDC) helps it meet local requirements for cloud environments and accelerate AI adoption
Long-term partners Orange and Google Cloud have expanded their partnership to deploy artificial intelligence, including gen AI, closer to Orange’s and its customers’ operations.
Orange said its new deployments aim to expand opportunities in countries and industries where local cloud environments are required. Google Distributed Cloud is a fully managed hardware and software solution that can be used to build and train AI models on Google Cloud using selected Vertex AI services and deploy and run locally at the edge. The telco can also use it to deploy ready to use Google Cloud products, such as Dataproc and AlloyDB as the building blocks for a custom-built solution.
Building applications with the latest in AI innovations can be challenging and scaling AI across multiple edge locations adds to the complexity. However, key to the service is that customers can build apps in full isolation with no connectivity to the public internet which is important for global telcos wanting to provide sovereign AI services.
This so-called air-gapped option does not require connectivity to Google Cloud or the public internet to manage the infrastructure, services, APIs, or tooling, and is built to remain disconnected in perpetuity. Google said its open architecture creates familiarity across both public and air-gapped private clouds.
Orange will use GDC in three areas. The first is to have an environment to run sensitive network data and AI workloads that some country regulators may require to be local or on-premises. The second is that the telco will be able to run gen AI models on-premises in an environment that is integrated into similar Vertex AI services on Google Cloud.
Lastly, beyond the contact centre, Orange has leveraged AI and Google Cloud technology to deliver personalised recommendations for relevant phones, plans, and services–improving customer lifetime value. GDC also allows gen AI-based speech recognition to occur in each Orange country, bringing these AI technologies even to countries without a Google Cloud region.
“We have a mission to accelerate value creation for Orange with every job, every network, and every customer experience super-powered by responsible AI. Orange sees enormous value in AI across every dimension of our business,” said Orange Group CEO Christel Heydemann.
“This partnership with Google Cloud and the cutting-edge solutions announced today are foundational to Orange achieving AI at scale and is a major step towards unlocking significant value from all of our data,” she added.
Bringing cloud to Orange’s data centres
The companies said the collaboration will bring the cloud to Orange’s own data centres, protecting sensitive workloads that must stay on-premises and enabling Orange to filter extremely high-volume data, such as over one petabyte a day of network telemetry. Orange said it will also enable local teams to deploy AI applications faster.
“Businesses are increasingly bringing gen AI solutions to the edge of the network to ensure better agility, responsiveness, and resilience,” said Google Cloud CEO Thomas Kurian. “Our partnership with Orange addresses that need, combining data, reliable infrastructure, and leading AI technologies to create new solutions to meet Orange’s global needs.”
Orange and Google Cloud have collaborated since 2020 when they signed a strategic partnership to accelerate the transformation of Orange’s IT infrastructure and the development of future cloud services, in particular edge computing. Over the past three years, Orange has migrated 13 petabytes of data to Google Cloud and has many strategic use cases in production across ten countries running on Google Cloud.
DSIT says this brings fibre broadband investment this year so far to £714m but it’s still only spent £1.3bn in total of the £5bn promised in 2019
The UK’s Department for Science, Technology & Innovation (DSIT) has announced a further £165 million to bring fibre broadband to almost 380,000 premises this year. They are in rural areas, including parts of Devon and the Isles of Scilly, South Yorkshire, the Peak District (pictured), Cornwall and Somerset.
The contracts have been awarded to:
South Yorkshire – Quickline awarded £44.4 million to connect 32,100 premises
West Herefordshire and the Forest of Dean – FullFibre awarded £23.4 million to connect 7,900 premises
Peak District – FullFibre awarded £10.7 million to connect 4,400 premises
Dorset and South Somerset –Wessex Internet awarded £33.5 million to connect 21,400 premises
Cornwall and the Isles of Scilly –Wildanet awarded £41.2 million to connect 16,800 premises
Mid-west Shropshire – Voneus awarded £12 million to connect 6,000 premises.
The investment is part of the UK Government’s is still calling its “£5 billion Project Gigabit programme” which was rebranded as such in March 2021. The plan was to bring fast broadband (up to 1Gbps) to hard to reach areas that are typically not densely populated. So far, the total invested is 1.3 billion.
However, before the General Election in 2019, the Tory government promised it would contribute £5 billion to get gigabit coverage to the 20% of the country’s population that are hard to reach, to ensure the whole country had gigabit access by 2025.
In October 2020, the government said it would release only £1.2 billion by 2025 and lowered the target to 85% coverage of the population, but said it was open to ideas from network providers to hand out more.
Almost 82% of properties across the UK now have access to fast broadband, up from 7% five years ago, according to the DSIT statement.
Not so fast
Although the Minister for Data and Digital Julia Lopez said, “The figures published today demonstrate how rapidly we are delivering higher quality gigabit broadband to every part of the country – even some of the most remote rural areas,” the government is not principally responsible for the dramatic acceleration in deployment.
Given that the intention was announced in 2019, getting started in 2022 was in fact slow, pandemic notwithstanding.
The Government has also awarded separate contracts to connect up to 800 primary schools in England, jointly funded through Project Gigabit and the Department for Education.
