Hosting Journalist

Hosting Journalist Hosting Journalist is a news portal covering the global web & cloud hosting industry. For your daily

Hostingjournalist.com is a news portal covering the global hosting industry with solutions among which cloud servers, dedicated servers virtual servers, reseller hosting, managed hosting, CDN, and colocation.

    Rad Web Hosting Now Offering VPS Hosting in France: This expansion brings powerful, secure, and scalable VPS hosting...
14/09/2025

Rad Web Hosting Now Offering VPS Hosting in France: This expansion brings powerful, secure, and scalable VPS hosting closer to European clients, reinforcing the company’s mission to deliver reliable cloud solutions worldwide. The new France-based VPS hosting option offers businesses, developers, and...

http://dlvr.it/TN4Mr1

    PsiQuantum Secures $1B to Build Fault-Tolerant Million-Qubit Computers: Quantum computing start-up PsiQuantum has se...
14/09/2025

PsiQuantum Secures $1B to Build Fault-Tolerant Million-Qubit Computers: Quantum computing start-up PsiQuantum has secured $1 billion in fresh funding to accelerate its ambitious plan to build the world’s first utility-scale, fault-tolerant quantum computers. The Series E round, announced this week, values the company at $7 billion and represents one of the largest single investments to date in quantum technology.

PsiQuantum says the capital will be used to establish sites in Brisbane, Australia, and Chicago, where large-scale systems can be deployed and validated, while advancing the performance of its quantum photonic chips and its error-corrected architecture.

The round was led by funds and accounts managed by affiliates of BlackRock, with participation from Temasek and Baillie Gifford. New investors include entities managed by Macquarie Capital, Ribbit Capital, NVentures - NVIDIA’s venture arm - Adage Capital Management, Qatar Investment Authority, Type One Ventures, Counterpoint Global (Morgan Stanley), 1789 Capital, and S Ventures (SentinelOne). Existing backers, including Blackbird, Third Point Ventures, and T. Rowe Price Associates, also participated.

Founded in 2016 by scientists Jeremy O’Brien and Pete Shadbolt, PsiQuantum has built its mission on the principle that commercially relevant quantum computing requires fault tolerance at the million-qubit scale. Current quantum prototypes across the industry typically operate with only a few hundred qubits and face challenges of scaling, stability, and manufacturability. PsiQuantum argues that photonic qubits - encoded in particles of light - offer a practical pathway, since they can be fabricated using established semiconductor processes.

Prof. Jeremy O’Brien, the company’s CEO, described the strategy as a deliberate choice to tackle engineering barriers head-on rather than treating the project as an open-ended scientific experiment. “Only building the real thing - million-qubit-scale, fault-tolerant machines - will unlock the promise of quantum computing,” he said. “We tackled the hardest problems first at the architectural and chip level, and are now mass-manufacturing quantum photonic chips at a leading U.S. semiconductor fab. With this funding, we’re ready to take the next decisive steps.”

Central to that progress is PsiQuantum’s integration of Barium Titanate (BTO), a high-performing electro-optic material, into its manufacturing process. BTO enables ultra-fast optical switches, considered a missing piece in the scalability of photonic quantum computing. PsiQuantum has already demonstrated the ability to produce 300-millimeter wafers of BTO in its California facilities, later combining them with wafers manufactured at GlobalFoundries. The company believes scaling BTO production is essential not only for quantum systems but also for next-generation AI supercomputers, which are increasingly demanding low-power, high-speed optical networking.

The photonic architecture sidesteps some of the most visible hardware limitations faced by competing quantum technologies. Unlike superconducting qubits, which require bulky cryostats resembling chandeliers, PsiQuantum has designed a modular high-density cooling solution resembling data center racks. Each cabinet can cool hundreds of quantum chips simultaneously. The company has also demonstrated high-fidelity networking between cabinets using standard telecom fiber, which is expected to be essential for distributing workloads in utility-scale machines.

Collaboration with NVIDIA

Pete Shadbolt, PsiQuantum’s Chief Scientific Officer and co-founder, said the company has reached a critical stage. “Nearly nine years after we started, we have pushed the technology to an unprecedented level of maturity and performance. We have the chips, we have the switches, we have scalable cooling technology, we can do networking, we have identified the sites, and we have both commercial and government support. We’re ready to get on and build utility-scale systems.”

The company’s collaboration with NVIDIA adds another dimension. Beyond its investment through NVentures, NVIDIA is working with PsiQuantum on areas such as quantum algorithms, GPU-quantum processor integration, and software frameworks tailored for silicon photonics. As hybrid quantum-classical computing becomes a near-term priority, integration between GPUs and QPUs is expected to be vital for enterprise and scientific workloads.

The Brisbane and Chicago sites will serve as the foundation for the first generation of PsiQuantum’s large-scale deployments, supported by government partnerships and local infrastructure initiatives. The company says these facilities will enable it to validate systems integration at scale, a prerequisite for commercializing machines capable of solving problems beyond the reach of classical supercomputers.

PsiQuantum has been scaling its manufacturing capabilities since its $450 million Series D in 2021. At that time, it announced a partnership with GlobalFoundries to fabricate integrated photonic chipsets containing the core elements of a quantum computer. These chips, designed by PsiQuantum, already exceed the performance of previously state-of-the-art photonic devices, according to the company. With the new financing, PsiQuantum says it can move from experimental prototypes toward production-grade systems that enterprises and governments can realistically deploy.

Quantum computing remains a field characterized by both vast potential and profound technical uncertainty. Proponents argue that fault-tolerant quantum machines could revolutionize sectors such as drug discovery, materials science, logistics, and cryptography by solving problems intractable for classical systems. However, the technical hurdles to achieve large-scale, error-corrected qubits have kept the field in the realm of research labs. PsiQuantum’s billion-dollar raise underscores both the appetite among investors for breakthrough technologies and the growing conviction that photonics may offer a faster route to scale than competing approaches.

As competition intensifies among quantum start-ups, tech giants, and national initiatives, PsiQuantum’s latest move positions it as one of the best-funded independent players in the race. With manufacturing partnerships in place, a billion-dollar war chest, and the backing of institutional investors and strategic partners, the company is now entering a decisive phase where promises of scale will be tested against the realities of engineering ex*****on.

For enterprises, the developments point to a potential timeline where commercially useful quantum computing may emerge sooner than many expected, though the exact timeframe remains uncertain. For PsiQuantum, the Series E funding signals confidence that after years of foundational work, the path toward a million-qubit machine is not just theoretical but increasingly tangible.

http://dlvr.it/TN4M9K

  EcoDataCenter Secures Additional EUR 600M Financing: EcoDataCenter has secured an additional EUR 600 million in debt f...
14/09/2025

EcoDataCenter Secures Additional EUR 600M Financing: EcoDataCenter has secured an additional EUR 600 million in debt financing from Deutsche Bank Private Credit and Infrastructure, marking another milestone in the company’s push to expand its high-performance data center footprint across Europe.

The Swedish operator, which positions itself at the intersection of AI infrastructure and sustainable digital services, plans to channel the new capital into the ongoing development of its campuses in Falun and Borlänge. The move comes as demand for advanced computing capacity continues to surge, particularly for AI-driven workloads.

EcoDataCenter has Chief executive Peter Michelson described AI infrastructure as a “new base industry,” underlining the company’s ambition to become a leading European player in high-performance computing. The financing follows a series of strategic steps that have established EcoDataCenter as one of Europe’s faster-growing digital infrastructure providers.

In 2024, the company partnered with U.S.-based AI hyperscaler CoreWeave to build one of the continent’s largest AI clusters in Falun, a project that signaled Sweden’s rising importance on the global digital infrastructure map. Soon after, EcoDataCenter purchased the former Kvarnsveden paper mill in Borlänge to convert the industrial site into additional data center capacity.

EUR 1.8B Financing Secured Since 2023

With its latest round, EcoDataCenter and its owner Areim have secured roughly EUR 1.8 billion in financing since 2023. Chief financial officer Johan Rydmark emphasized that attracting capital of this magnitude highlights both investor confidence and the scalability of the company’s platform. Clients already include global enterprises such as BMW, translation service DeepL, and CoreWeave, further reinforcing EcoDataCenter’s credibility as a partner for compute-intensive workloads.

Founded in 2019 with its first facility in Falun, EcoDataCenter has built its brand around technological leadership and a sustainability-driven approach. The company’s expansion strategy reflects the broader industry trend where AI and high-performance computing are reshaping the economics of digital infrastructure.

LionTree Advisors acted as financial advisor on the transaction, while White & Case LLP provided legal counsel.

http://dlvr.it/TN4LSv

    Digital Realty Opens Innovation Lab for AI and Hybrid Cloud Testing: Digital Realty, one of the world’s largest prov...
14/09/2025

Digital Realty Opens Innovation Lab for AI and Hybrid Cloud Testing: Digital Realty, one of the world’s largest providers of cloud- and carrier-neutral data centers, has unveiled a new initiative designed to help enterprises experiment with artificial intelligence (AI) and hybrid cloud infrastructure in real-world conditions before going to market.

The company this week announced the launch of the Digital Realty Innovation Lab (DRIL), a dedicated environment within its Northern Virginia campus where partners and customers can test high-density workloads, validate AI deployments, and refine hybrid strategies without risking disruption to production systems.

The facility, situated in a live colocation data center, enables enterprises to bring their own workloads or work with pre-configured infrastructure provided by Digital Realty and its partners. By replicating production-grade architectures in a contained environment, the lab aims to remove a longstanding barrier in enterprise IT: the lack of practical, large-scale testing grounds where new infrastructure can be validated under real conditions. Customers are able to measure performance, tune configurations, and connect seamlessly with cloud and network providers through ServiceFabric, Digital Realty’s global orchestration and interconnection platform.

Chris Sharp, Digital Realty’s Chief Technology Officer, emphasized the business urgency behind the launch. “Innovation isn’t optional – it’s a competitive requirement in a world where data volumes and AI adoption are accelerating,” he said. “Yet, many enterprises still deploy complex infrastructure without the ability to test, validate, or optimize in real-world conditions. The DRIL changes that. It provides a live, high-density environment where customers gain real-time, data-driven insights to fine-tune their deployments before scaling.”

The new lab has been designed with several enterprise priorities in mind. It accommodates power-hungry AI and high-performance computing (HPC) workloads, supporting up to 150 kilowatts per cabinet. It also integrates with the ePlus AI Experience Center, enabling businesses to explore AI-specific requirements around cooling, power, and GPU resources. With direct cloud connectivity, companies can refine hybrid strategies and onboard with ServiceFabric via cross connects, while the Private AI Exchange (AIPx) allows them to orchestrate workloads across different environments. Latency testing across multiple sites is also supported, allowing enterprises to simulate global-scale deployments.

The announcement has drawn strong support from hardware and infrastructure partners that see the lab as an accelerator for enterprise AI adoption. AMD highlighted its role in showcasing the performance of its EPYC processors and Instinct accelerators in a high-density, liquid-cooled environment. Robert Hormuth, Corporate Vice President for Architecture and Strategy at AMD, said the lab represents “a production-grade opportunity to validate real-world use cases and help organizations bring AI-powered solutions to market faster.”

Cisco echoed that sentiment, with Kevin Wollenweber, Senior Vice President and General Manager for Cisco’s Data Center and Internet Infrastructure, underscoring the risks of AI deployments without validation. “AI investment requires complete confidence. The infrastructure supporting AI workloads is too critical to blindly trust without testing,” he said, noting that Cisco’s AI Pods would be available within the lab to test networking, compute, and security requirements across different AI use cases.

CommScope described the project as a showcase of intelligent physical design. The company is providing the fiber and cabling infrastructure underpinning the lab’s rapid deployment capabilities. Lenovo pointed to its Neptune liquid-cooling systems, which are being integrated to support high-performance private AI workloads. Supermicro highlighted the importance of validation, with Chief Growth Officer Cenly Chen noting that customers could ensure optimal performance of products on Supermicro’s advanced platforms before deployment.

Other partners emphasized the lab’s role in unlocking data mobility and extending global reach. Vcinity’s CEO Harry Carr said the initiative will enable enterprises to “access and act on data wherever it resides” and monetize AI investments more quickly. Zenlayer, which is using the lab to showcase AI and edge applications, described it as a way to help enterprises build future-proof expansion strategies across emerging markets such as Asia-Pacific, Latin America, the Middle East, and Africa.

The launch of the DRIL builds on Digital Realty’s earlier AI Experience Center, opened in 2024 with ePlus and Vertiv, which was designed to explore AI-centric use cases. The new lab extends that concept by embedding AI testing capabilities directly within an operational data center, providing what Ken Farber, President of ePlus Software, described as “a real-world proving ground that combines strategy, infrastructure design, and global connectivity to turn ideas into AI solutions.”

The Northern Virginia site is the first to host the DRIL, but Digital Realty has already confirmed expansion plans. London will become the next location in early 2026, with further rollouts expected across other global campuses. By connecting sites, the company will enable customers to simulate latency scenarios between geographies, further supporting multinational enterprises in validating global cloud and AI strategies.

For Digital Realty, the initiative underscores its ambition to position colocation not just as a commodity infrastructure service, but as a catalyst for enterprise transformation in an AI-driven economy. As CTO Chris Sharp put it, “Through initiatives like the DRIL and strategic partnerships, Digital Realty is enabling enterprises to accelerate innovation, optimize performance, and drive a more connected, intelligent digital future.”

http://dlvr.it/TN4L3y

    IDC: Global Semiconductor Revenue to Hit $800B in 2025 on AI Demand: The global semiconductor industry is heading in...
13/09/2025

IDC: Global Semiconductor Revenue to Hit $800B in 2025 on AI Demand: The global semiconductor industry is heading into another record year, with revenue projected to hit $800 billion in 2025, according to new data from International Data Corporation (IDC). That figure represents a 17.6% increase from the $680 billion forecast for 2024, itself a year of strong recovery with 22.4% growth after a period of contraction.

IDC attributes the expansion primarily to data center demand, particularly the surge in AI infrastructure, accelerated computing, and networking technologies.

The compute segment of the semiconductor market is expected to grow 36% year-over-year to $349 billion in 2025, driven by hyperscale AI workloads. IDC projects this segment to sustain a compound annual growth rate of 12% through 2030. The acceleration is so pronounced that, for the first time, a single semiconductor vendor is expected to surpass $200 billion in annual revenue, underscoring the dominance of AI-driven data center spending. Adjacent markets such as rack-scale systems, high-speed interconnects, memory, and networking semiconductors are also benefiting from this momentum.

Networking is emerging as a critical bottleneck in AI workloads, with data movement increasingly constraining system performance more than compute power itself. To address this, cloud providers, telecommunications firms, and enterprises are ramping up investment in networking chips and optical interconnects. IDC forecasts networking and connectivity semiconductor demand to grow 13% in 2025, led by high-capacity Ethernet switches, SmartNICs, and DPUs, which help offload network processing from CPUs and GPUs. Optical technologies are also gaining traction as data-intensive AI applications push traditional copper-based systems to their limits.

Beyond data centers, other markets are beginning to rebound after periods of weakness. Automotive and industrial semiconductors, which suffered from excess inventory and demand softness in 2024, are now stabilizing. Automotive suppliers are reporting sequential growth as inventory levels normalize, though headwinds remain from expiring subsidies in China, ongoing price pressure, and trade uncertainty. Still, rising semiconductor content per vehicle, driven by electrification, silicon carbide (SiC) and gallium nitride (GaN) adoption, and the shift toward domain and zonal controllers, will sustain growth. IDC expects the automotive semiconductor sector to increase by 3% in 2025.

Improved Visibility and Recovery

IDC’s industrial semiconductor market is on a stronger recovery path, with 11% growth expected in 2025 after contracting nearly 14% the previous year. Demand is being bolstered by military and aerospace applications, manufacturing, edge AI adoption, and electrification trends. While macroeconomic uncertainty and cautious capital spending remain challenges, industrial chip suppliers are reporting improved visibility and backlog recovery.

Smartphone semiconductors, while no longer driving growth through unit volume, are seeing higher revenue per device. The wireless semiconductor segment is forecast to grow 5% in 2025, with adoption of 5G, richer multimedia capabilities, and on-device AI driving higher content. Average selling prices are climbing as smartphones integrate neural processing units (NPUs), GPUs, and advanced connectivity features. However, tariff policies and trade restrictions may distort shipment timing and affect consumer pricing into 2026.

Industry analysts emphasize that while markets like automotive and industrial are only beginning to recover, their long-term trajectory remains strong due to increased chip content and electrification. “While these markets are not experiencing the explosive growth of the datacenter, the increase in semiconductor content per system, increased compute capabilities, and electrification will help ensure long-term revenue resilience,” said Nina Turner, research director at IDC.

Mario Morales, group vice president for semiconductors at IDC, noted that the industry is experiencing a structural shift: “Explosive demand for compute and networking at scale is creating a step-function in revenue growth, while adjacent markets from cloud to connectivity benefit from the shift to rack-scale systems. The overall industry is set on a strong trajectory that extends well beyond 2025.”

IDC now forecasts that the semiconductor market will surpass $1 trillion in revenue by 2028, nearly two years earlier than previous consensus estimates, cementing its place as one of the fastest-growing and most strategically critical sectors in the global economy.

http://dlvr.it/TN3gzd

    Supermicro Rolls Out NVIDIA Blackwell Ultra Systems for AI Factories: Supermicro has begun the global rollout of its...
13/09/2025

Supermicro Rolls Out NVIDIA Blackwell Ultra Systems for AI Factories: Supermicro has begun the global rollout of its NVIDIA Blackwell Ultra solutions, marking a significant milestone in the company’s long-standing collaboration with NVIDIA. The announcement centers on the broad availability of Plug-and-Play (PnP)-ready NVIDIA HGX B300 systems and GB300 NVL72 racks, pre-engineered and validated at system, rack, and even full data center scale.

These turnkey solutions are targeted at enterprises building the next generation of AI factories, where massive computational performance and efficiency are critical to training, inference, and deployment of advanced artificial intelligence models.

Charles Liang, president and CEO of Supermicro, highlighted the importance of these developments for data center operators struggling with the complexity of scaling AI infrastructure. “Supermicro has the best track record of fast and successful deployments of new NVIDIA technologies,” he said, adding that the company’s Data Center Building Block Solutions and on-site deployment expertise make it possible to deliver “the highest-performance AI platform” with reduced time-to-market. According to Liang, the challenges of cabling, power distribution, and cooling are growing as AI clusters expand, and pre-validated, plug-and-play systems represent a crucial advantage for organizations racing to deploy large-scale AI capacity.

The Blackwell Ultra platform brings significant generational advancements. At the system level, Supermicro integrates advanced air and direct liquid cooling designs optimized for GPUs capable of consuming up to 1,400 watts each. Compared to the previous Blackwell generation, the Ultra version delivers 50% greater inferencing performance using FP4 compute, alongside 50% more HBM3e memory capacity. These enhancements are aimed at running larger models with higher efficiency, a necessity as AI workloads push into trillions of parameters.

Supermicro’s NVIDIA Blackwell Ultra portfolio spans a wide range of configurations, including rack-scale GB300 NVL72 systems and 8U air-cooled or 4U liquid-cooled HGX B300 servers. The GB300 NVL72 rack-scale platform alone delivers 1.1 exaFLOPS of dense FP4 compute performance. Meanwhile, individual B300 systems provide up to 144 petaFLOPS of FP4 performance and 270 GB of HBM3e memory per GPU, representing up to 7.5x the performance of systems based on NVIDIA’s Hopper architecture.



Networking forms a key part of these solutions. Enterprises can choose between NVIDIA Quantum-X800 InfiniBand and Spectrum-X Ethernet fabrics, each offering bandwidth up to 800 Gb/s, ensuring Blackwell Ultra systems can be interconnected into large-scale clusters without bottlenecks. Integration of NVIDIA’s ConnectX-8 SuperNICs further boosts network throughput, doubling compute network bandwidth compared to earlier platforms.

NVIDIA AI Enterprise Software

Supermicro positions these systems not only as hardware but as part of a full-stack ecosystem. Each deployment is integrated with NVIDIA AI Enterprise software, along with blueprints and NIM microservices designed to accelerate AI workloads. Beyond hardware and software, the company also provides extensive deployment services through its Data Center Building Block Solutions (DCBBS). This includes on-site cabling, thermal management, and power integration, ensuring faster time-to-online for customers building AI factories.

Sustainability is also an integral part of the pitch. Supermicro claims that its DLC-2 liquid cooling technology can save up to 40% in power consumption, reduce data center footprint by 60%, and cut water usage by 40%. Collectively, these improvements can lower total cost of ownership by as much as 20%, a key factor for enterprises facing mounting energy costs and sustainability requirements.

The availability of Blackwell Ultra solutions underscores the rapid evolution of infrastructure designed for AI. With pre-validated, scalable systems capable of exascale performance, Supermicro aims to simplify the complexity of deploying AI at scale, while enabling enterprises to future-proof their operations for the demands of multimodal models, agentic AI, and real-time inference.

http://dlvr.it/TN3gjK

  Rad Web Hosting Now Offering VPS Hosting in France: This expansion brings powerful, secure, and scalable VPS hosting c...
13/09/2025

Rad Web Hosting Now Offering VPS Hosting in France: This expansion brings powerful, secure, and scalable VPS hosting closer to European clients, reinforcing the company’s mission to deliver reliable cloud solutions worldwide. The new France-based VPS hosting option offers businesses, developers, and...

http://dlvr.it/TN3gBv

    Hostinger’s AI Assistant Kodee Saves €9M Yearly, Boosts Efficiency: Hostinger, one of the fastest-growing hosting pr...
13/09/2025

Hostinger’s AI Assistant Kodee Saves €9M Yearly, Boosts Efficiency: Hostinger, one of the fastest-growing hosting providers worldwide, has significantly expanded the capabilities of its AI-powered chat assistant, Kodee. The virtual assistant, which now performs more than 350 distinct tasks for customers, would save the company more than €750,000 each month - amounting to annualized savings of over €9 million - while also driving measurable improvements in client satisfaction.

What began as a tool for answering queries has evolved into a proactive AI agent that directly executes tasks such as server security, website migration, speed optimization, and domain management.

Giedrius Zakaitis, Chief Product and Technology Officer at Hostinger, said that the impact has been immediate. “Kodee conducted 750,000 conversations in August alone, resolving 75% of them without any human intervention. That’s up from 50% at the beginning of 2025. When Kodee acts instead of only answering, resolution rates and customer satisfaction improve by almost four percentage points,” he explained.

Hostinger attributes Kodee’s success to three core factors: speed, scalability, and cost efficiency. On speed, the AI now responds in seconds, slashing average wait times for assistance from 28 seconds a year ago to just 9 seconds today. The company reports that 93% of all queries are resolved in under two minutes, compared to 77% one year earlier.

On scalability, the company is handling more than double the number of client inquiries compared to four years ago - nearly matching 2024’s full-year total in just eight months of 2025 - without needing to hire at the same rate.

The customer success team has been reduced to 312 specialists, down 20% from earlier this year, as many routine roles are now automated. Zakaitis emphasized that no jobs were lost to AI; rather, roles were redefined, enabling human staff to focus on complex cases.

The financial impact is equally striking. To process today’s volume of inquiries without AI, Hostinger estimates it would need an additional 700 to 800 human agents. Instead, a team of just 14 people manages and enhances Kodee. Beyond cost savings, the AI also reduces reliance on multilingual support staff by conversing fluently in over 50 languages.



Open-Source Model Context Protocol

Built using the open-source Model Context Protocol (MCP), Kodee connects directly to Hostinger’s infrastructure. Its capabilities span the company’s entire service portfolio: managing DNS records, configuring domains, checking server health, running malware scans, handling payments, writing and optimizing blog content, managing WooCommerce products, and even editing images within Hostinger’s website builder. The assistant can also act as a first-line operations agent for VPS customers, running virus scans and granting access requests.

Customer behavior suggests that Kodee is perceived as more than just software. “One client added a Google verification code without explanation, and Kodee immediately recognized the need for a DNS record, requested approval, and executed it. Some customers even thank Kodee personally and call it ‘bro,’ as if they’re talking to a human,” Zakaitis noted.

Kodee sits within a broader AI strategy at Hostinger. Earlier this year, the company launched Hostinger Horizons, an AI-powered website and web app builder, and Hostinger Reach, an AI-based email marketing tool. The Hostinger Website Builder also leverages AI to automate text generation, graphic creation, SEO management, and design optimization. Together, these offerings aim to accelerate website creation and digital growth for the company’s more than 4 million customers worldwide.

By reducing support costs, accelerating response times, and helping users manage increasingly complex workloads, Kodee would represent more than an operational efficiency tool. It is reshaping how hosting firms can scale in an AI-first era. With adoption still growing and new features being added regularly, Hostinger appears intent on setting a new standard for AI-driven customer service and digital infrastructure support.

http://dlvr.it/TN3g3r

  PsiQuantum Secures $1B to Build Fault-Tolerant Million-Qubit Computers: Quantum computing start-up PsiQuantum has secu...
13/09/2025

PsiQuantum Secures $1B to Build Fault-Tolerant Million-Qubit Computers: Quantum computing start-up PsiQuantum has secured $1 billion in fresh funding to accelerate its ambitious plan to build the world’s first utility-scale, fault-tolerant quantum computers. The Series E round, announced this week, values the company at $7 billion and represents one of the largest single investments to date in quantum technology.

PsiQuantum says the capital will be used to establish sites in Brisbane, Australia, and Chicago, where large-scale systems can be deployed and validated, while advancing the performance of its quantum photonic chips and its error-corrected architecture.

The round was led by funds and accounts managed by affiliates of BlackRock, with participation from Temasek and Baillie Gifford. New investors include entities managed by Macquarie Capital, Ribbit Capital, NVentures - NVIDIA’s venture arm - Adage Capital Management, Qatar Investment Authority, Type One Ventures, Counterpoint Global (Morgan Stanley), 1789 Capital, and S Ventures (SentinelOne). Existing backers, including Blackbird, Third Point Ventures, and T. Rowe Price Associates, also participated.

Founded in 2016 by scientists Jeremy O’Brien and Pete Shadbolt, PsiQuantum has built its mission on the principle that commercially relevant quantum computing requires fault tolerance at the million-qubit scale. Current quantum prototypes across the industry typically operate with only a few hundred qubits and face challenges of scaling, stability, and manufacturability. PsiQuantum argues that photonic qubits - encoded in particles of light - offer a practical pathway, since they can be fabricated using established semiconductor processes.

Prof. Jeremy O’Brien, the company’s CEO, described the strategy as a deliberate choice to tackle engineering barriers head-on rather than treating the project as an open-ended scientific experiment. “Only building the real thing - million-qubit-scale, fault-tolerant machines - will unlock the promise of quantum computing,” he said. “We tackled the hardest problems first at the architectural and chip level, and are now mass-manufacturing quantum photonic chips at a leading U.S. semiconductor fab. With this funding, we’re ready to take the next decisive steps.”

Central to that progress is PsiQuantum’s integration of Barium Titanate (BTO), a high-performing electro-optic material, into its manufacturing process. BTO enables ultra-fast optical switches, considered a missing piece in the scalability of photonic quantum computing. PsiQuantum has already demonstrated the ability to produce 300-millimeter wafers of BTO in its California facilities, later combining them with wafers manufactured at GlobalFoundries. The company believes scaling BTO production is essential not only for quantum systems but also for next-generation AI supercomputers, which are increasingly demanding low-power, high-speed optical networking.

The photonic architecture sidesteps some of the most visible hardware limitations faced by competing quantum technologies. Unlike superconducting qubits, which require bulky cryostats resembling chandeliers, PsiQuantum has designed a modular high-density cooling solution resembling data center racks. Each cabinet can cool hundreds of quantum chips simultaneously. The company has also demonstrated high-fidelity networking between cabinets using standard telecom fiber, which is expected to be essential for distributing workloads in utility-scale machines.

Collaboration with NVIDIA

Pete Shadbolt, PsiQuantum’s Chief Scientific Officer and co-founder, said the company has reached a critical stage. “Nearly nine years after we started, we have pushed the technology to an unprecedented level of maturity and performance. We have the chips, we have the switches, we have scalable cooling technology, we can do networking, we have identified the sites, and we have both commercial and government support. We’re ready to get on and build utility-scale systems.”

The company’s collaboration with NVIDIA adds another dimension. Beyond its investment through NVentures, NVIDIA is working with PsiQuantum on areas such as quantum algorithms, GPU-quantum processor integration, and software frameworks tailored for silicon photonics. As hybrid quantum-classical computing becomes a near-term priority, integration between GPUs and QPUs is expected to be vital for enterprise and scientific workloads.

The Brisbane and Chicago sites will serve as the foundation for the first generation of PsiQuantum’s large-scale deployments, supported by government partnerships and local infrastructure initiatives. The company says these facilities will enable it to validate systems integration at scale, a prerequisite for commercializing machines capable of solving problems beyond the reach of classical supercomputers.

PsiQuantum has been scaling its manufacturing capabilities since its $450 million Series D in 2021. At that time, it announced a partnership with GlobalFoundries to fabricate integrated photonic chipsets containing the core elements of a quantum computer. These chips, designed by PsiQuantum, already exceed the performance of previously state-of-the-art photonic devices, according to the company. With the new financing, PsiQuantum says it can move from experimental prototypes toward production-grade systems that enterprises and governments can realistically deploy.

Quantum computing remains a field characterized by both vast potential and profound technical uncertainty. Proponents argue that fault-tolerant quantum machines could revolutionize sectors such as drug discovery, materials science, logistics, and cryptography by solving problems intractable for classical systems. However, the technical hurdles to achieve large-scale, error-corrected qubits have kept the field in the realm of research labs. PsiQuantum’s billion-dollar raise underscores both the appetite among investors for breakthrough technologies and the growing conviction that photonics may offer a faster route to scale than competing approaches.

As competition intensifies among quantum start-ups, tech giants, and national initiatives, PsiQuantum’s latest move positions it as one of the best-funded independent players in the race. With manufacturing partnerships in place, a billion-dollar war chest, and the backing of institutional investors and strategic partners, the company is now entering a decisive phase where promises of scale will be tested against the realities of engineering ex*****on.

For enterprises, the developments point to a potential timeline where commercially useful quantum computing may emerge sooner than many expected, though the exact timeframe remains uncertain. For PsiQuantum, the Series E funding signals confidence that after years of foundational work, the path toward a million-qubit machine is not just theoretical but increasingly tangible.

http://dlvr.it/TN3fM4

Adres

Coehoornsingel 58
Zutphen
7201AD

Meldingen

Wees de eerste die het weet en laat ons u een e-mail sturen wanneer Hosting Journalist nieuws en promoties plaatst. Uw e-mailadres wordt niet voor andere doeleinden gebruikt en u kunt zich op elk gewenst moment afmelden.

Contact

Stuur een bericht naar Hosting Journalist:

Delen