Mount Thor raised $45 million to expand Apple Silicon-based cloud infrastructure for AI
Read more
Ventureburn
ventureburn.com

Mount Thor raised $45 million to expand Apple Silicon-based cloud infrastructure for AI

Mount Thor, an artificial intelligence cloud infrastructure provider, announced the raising of $45 million in a joint Seed and Series A funding round. The company offers a managed platform for utilizing Apple hardware in AI-related tasks.

The Series A funding was co-led by CRV and Conviction. The Seed round was led by Chad Byers and Offline Ventures. Other investors included CSC Leasing, Enea Capital, Long Journey Ventures, and Upper90. Several strategic private investors also participated in both rounds.

Investors included companies associated with Cognition, Core Automation, Daytona, Meta, and OpenAI. Some investors from Together AI and Vercel also participated in the financing. Additionally, Mount Thor received non-dilutive funding to acquire hardware.

The company's Voell platform integrates Apple hardware with cloud infrastructure. The system develops, operates, and manages rack-scale systems within data centers, providing clients with on-demand access to computing resources for remote access to macOS and Apple Silicon.

This infrastructure allows researchers to work on AI, test software, and perform model inference. It also enables developers to create secure environments for autonomous agents running on macOS, while research groups can use the hardware for reinforcement learning environments.

As autonomous agents increasingly perform complex tasks by interacting with computers, the need for computing environments that simplify the use of established software ecosystems is growing. Mount Thor considers macOS a critical component for such agents, and its infrastructure allows clients to use Apple environments remotely.

The company has also deployed for inference workloads in the enterprise sector and application simulation. The raised funds will allow the firm to scale its computing power and hire more personnel. One priority area will be the creation of an inference lab for open-source models specifically adapted for Apple Silicon to meet demand for local and cloud workloads.

Mount Thor aims to offer enterprises and AI developers a different type of inference infrastructure that can support workloads unsuitable for typical GPU-based cloud environments. The startup launched its first data center last week, completing deployment in less than six months. Its elastic Apple Silicon fleet is located with Tier 1 colocation service providers in all major US metropolitan areas. The Mount Thor team also has experience working in traditional computing infrastructure companies.

The fundraising also reflects the growing focus on infrastructure for autonomous AI systems. Mount Thor promotes Apple hardware as a solution for a range of tasks, offering services to research organizations, corporate clients, and applied AI firms, including secure agent sandboxes where developers can evaluate agent capabilities on macOS, and software developers can use remote machines to build and emulate Apple applications.

CRV invests in the company as the use of Apple platforms expands. CRV General Partner, Brittany Walker, noted that she observes a convergence between autonomous agents and increasing computational demands, and Mount Thor's success will depend on its ability to scale demand to utilize the infrastructure.

Similar stories

AI Startup Humyn Labs Raises $47 Million in Series A Round to Expand Global Operations
Read more
business-standard.com

AI Startup Humyn Labs Raises $47 Million in Series A Round to Expand Global Operations

Humyn Labs, a startup focused on researching and implementing physical artificial intelligence (AI), has raised $47 million in a Series A funding round. These funds are designated for expanding the company's global activities and building infrastructure for testing and deploying robots in real-world conditions, the firm announced on Wednesday.

The round includes $26 million in equity capital, led by Accel, with participation from Prosus Ventures, BeeGlobal, Finserv Intelligence through Bajaj Finance, Bajaj Alts AI Opportunities Fund, Dallas Venture Capital, angel investor Anita Pramoda, and Jump Trading LLC. The remaining $21 million was raised through venture debt, curated by Lighthouse Canton, with participation from Trifecta Capital.

In April of this year, Humyn Labs announced its intention to allocate $20 million to develop human data infrastructure for physical AI firms worldwide.

Plans for Utilizing New Capital

In an interview with Business Standard, co-founder Manish Agarwal stated that the startup plans to direct the fresh capital towards creating deployment-ready infrastructure in Brazil, South Korea, and India. Furthermore, the company plans to expand its capabilities in collecting and enriching hardware data, as well as strengthen its presence in San Francisco, London, and Seoul.

Founded in 2026 by Agarwal and Ishank Gupta, Humyn Labs achieved an annual revenue run rate exceeding $50 million in just six months of operation. The company collaborates with leading AI laboratories, robotics companies, and world model developers, collecting and processing data across more than 20 countries in Latin America, East Asia, and South and Southeast Asia.

Unlike traditional AI systems, which primarily rely on digital datasets, physical AI models require understanding and interaction with the real world. Humyn Labs captures how people perform tasks in factories, warehouses, farms, and homes, transforming this information into training-ready data covering vision, sound, motion, and touch.

However, Agarwal emphasized that the company's business goes beyond simple data collection. A more significant opportunity lies in helping robotics companies bridge the gap between model training and their actual deployment. He noted: 'Data collection is only the input side. The reason data is collected is the desire to deploy robotic forms in various use cases. Providing data alone is not enough for deployment readiness. You need to start deploying these models and hardware in the real world.'

The company plans to create specialized robot deployment centers, which Agarwal termed 'robot gyms,' where models and hardware can be tested in real-world conditions. Collaboration with large enterprises to establish similar centers at their facilities globally is also planned.

Although Humyn Labs is headquartered in Bengaluru, it has a subsidiary in the US where about 95 percent of its client base is concentrated. The company prioritizes San Francisco as its first international market, followed by London (UK), Munich (Germany), Seoul (South Korea), and Tokyo (Japan).

Despite being based in India, Humyn Labs does not consider the country an immediate priority for commercial robot deployment. Agarwal explained that deployment readiness depends on two factors: the economic viability of the robot and its ability to reliably perform tasks in real-world conditions. He added that India might become a priority market later, as the economics of robot deployment may not yet work in many use cases.

Corma raises $125 million at $1 billion valuation to scale AI in cybersecurity
Read more
ventureburn.com

Corma raises $125 million at $1 billion valuation to scale AI in cybersecurity

Corma, an Israeli cybersecurity startup, announced it has raised $125 million in a Series A funding round with the company valued at one billion dollars. This information was disclosed on November 16, 2026.

The funding was led by NEA, with participation from existing investors including Sequoia Capital, Khosla Ventures, and Coatue. This new capital came after the company previously announced a $60 million seed round in August 2026. Thus, over three months, Corma has raised a total of $185 million.

The company plans to use the funds to develop its foundational models and computing infrastructure. Alon Pluda, CEO and co-founder of Corma, noted that the situation represents a 'battle of wits' as more large organizations demand more advanced protection. In response, the company has increased its server capacity by 24 times.

Corma develops foundational models in defensive cybersecurity. Its artificial intelligence plays a central role in the operations of modern enterprise security teams. It is important to note that the product does not replace existing security products but works alongside them, allowing organizations to maintain their current level of cybersecurity.

The company's models process vast amounts of security data, such as audit logs, security events, and network traffic. They are capable of detecting unusual behavior across various systems and correlating small signals received over long periods, enabling security teams to investigate threats more effectively.

Corma asserts that coding-oriented models are incapable of handling such tasks. Defensive cybersecurity requires continuous analysis and making immutable decisions, as well as working with contradictory sources of information in complex corporate environments.

The company's tools perform security assurance functions. Their models are trained using reinforcement learning on simulated corporate networks. These simulated networks are designed to resemble security tools, telemetry, and operational noise.

There is reported growth in application deployments within companies listed in the Fortune 100 and Fortune 500, as well as interest from companies in the Fortune Global 500. These organizations operate across various industries, including healthcare, financial services, energy, and telecommunications.

Corma's move followed cybersecurity incidents in October, which were dubbed 'Black October.' During this period, numerous errors and breaches affected popular enterprise software, exposing weaknesses in corporate security systems.

Aaron Jacobson, a partner at NEA, stated that Corma successfully combines cutting-edge AI research with cybersecurity expertise. He emphasized that the firm's technology can help organizations prevent devastating cyberattacks, and the new funding round will provide it with sufficient capital to do so.

The Corma team includes scientists from Google DeepMind and cybersecurity experts from Israel's Unit 8200. The company has offices in Tel Aviv and San Francisco, and is led by management with experience in both offensive and defensive security.

The additional capital will help Corma further scale its proprietary models. The company intends to invest in computing infrastructure and hire a new group of researchers and engineers, although hiring plans remain moderate. Corma is developing a direction toward creating autonomous cybersecurity tools.

Other companies focused on autonomous security operations platforms include 7AI, Torq, and Exaforce. All their developments focus on automating security operations using AI agents. Corma's success will depend both on adoption in large enterprises and the availability of computational resources, as it needs to demonstrate effectiveness in various complex security scenarios.

Q-CTRL raises $290 million to scale AI-based quantum technologies
Read more
ventureburn.com

Q-CTRL raises $290 million to scale AI-based quantum technologies

The Australian quantum technology firm Q-CTRL has successfully raised $290 million USD in funding, positioning it as one of the new unicorns in the sector. The company described this funding round as the largest in the history of software-oriented quantum technology enterprises.

The round was extended for a second time in response to growing investor demand. New and returning investors from private equity (PE), venture capital (VC), and technology sectors participated. New participants include AMD Ventures, EDGE Group, MPower Partners, and Australia’s National Reconstruction Fund Corporation.

Existing investors, alongside the new ones, include Bullhound Capital, DCVC, In-Q-Tel, Lockheed Martin Ventures, and Square Peg. The funds raised will be used to deepen the company's software capabilities and facilitate its global expansion.

Q-CTRL's hardware-agnostic software platform utilizes artificial intelligence to enhance the performance of quantum computers and sensors. Its Fire Opal product helps minimize errors in quantum algorithms. The software suite includes enhancements for quantum computing, such as tools for assessing quantum computer performance. Researchers can use Boulder Opal to assist with qubit calibration, which contributes to improving the stability and practicality of quantum machines.

Furthermore, the company demonstrated material simulations on IBM quantum hardware. Developers have the ability to run their simulations on the IBM Quantum platform. Salesforce Ventures noted this achievement when explaining its investment decision.

Robert Kate, a partner at Salesforce Ventures, stated: 'The hardest part in quantum is not building the machine. It is extracting actionable results from it.' Q-CTRL's software is capable of integrating quantum technologies into corporate data centers, making them accessible to a wider range of users and applications.

Q-CTRL is also commercializing quantum sensing through its Ironstone Opal navigation systems. These systems are designed to provide navigation in the absence of GPS signals. The company's next step is planned international growth and scaling of operations.

Ironstone Opal uses magnetometers (quantum sensors) to measure variations in Earth's magnetic and gravitational fields. The resulting data allows for vehicle localization. This sensor data is supplemented with mapping data via advanced software. Q-CTRL asserts that its technology is resilient to both interference and spoofing. The product is already being installed on commercial aircraft and autonomous vehicles. The company is also collaborating with Lockheed Martin on unmanned aerial vehicle development.

It is reported that this is the first quantum navigation technology to meet flight safety requirements. CEO Michael J. Biercuk previously stated that the technology can determine its position with an error margin of about 10 meters, indicating the technology's transition out of laboratory settings.

The capital raise also signifies the addition of strategic connections through the expanded investor syndicate. Q-CTRL anticipates that these partnerships will unlock opportunities through joint sales and integrated technologies. Investors may also help create direct revenue streams in the sectors served by Q-CTRL.

Bullhound Capital highly praised the company's client relationships and revenue growth, noting that Q-CTRL skillfully commercializes research and development. The reinvestment demonstrates confidence in the company's ability to execute its plans. Square Peg highlighted Q-CTRL's work in aviation and autonomous navigation, while MPower Partners saw additional prospects for the company in Japan, mentioning industrial applications and government support for quantum commercialization.

Popular