Overview
CoreWeave, Inc. is an American cloud-computing startup specializing in GPU infrastructure for artificial intelligence (AI), machine learning (ML), and other compute-intensive workloads. Founded in 2017, the company has rapidly evolved from a cryptocurrency mining operation to a leading provider of cloud-based GPU resources.
Founding and Evolution
Initially established as Atlantic Crypto, CoreWeave was founded by Michael Intrator, Brian Venturo, and Brannin McBee. Following the 2018 cryptocurrency crash, the company pivoted its focus to providing GPU computing resources across various industries, rebranding as CoreWeave in 2019.
Infrastructure and Technology
CoreWeave's infrastructure is built on a Kubernetes-native architecture, offering:
- A broad range of NVIDIA GPUs, including H100, A100, and V100 models
- Bare metal servers without hypervisors for dedicated resources
- Cloud Native Networking and InfiniBand for high-speed, low-latency connections
- High-performance NVMe File System Volumes and accelerated Object Storage
Key Use Cases
CoreWeave supports diverse intensive workloads, including:
- AI and Machine Learning: Model training and inference
- VFX and Rendering: Cloud rendering and virtual workstations
- Life Sciences: Drug discovery and molecular dynamics simulations
- Metaverse and Pixel Streaming: Interactive 3D content delivery
Funding and Valuation
CoreWeave has secured significant funding, including:
- $1.1 billion raised in May 2024, valuing the company at $19 billion
- $2.3 billion debt financing facility in August 2023
- Additional investments bringing the valuation to $23 billion by October 2024
Global Presence and Expansion
As of 2024, CoreWeave operates 13 data centers in the United States and two in the United Kingdom. The company plans further expansion, including a $1.2 billion data center in New Jersey and significant investments in the U.K.
Leadership and Future Plans
Led by CEO Michael Intrator, CTO Brian Venturo, and CSO Brannin McBee, CoreWeave employs around 550 people. The company plans to launch an initial public offering (IPO) in 2025 and continues to expand its AI infrastructure and global operations.
Leadership Team
CoreWeave's leadership team comprises experienced executives driving the company's strategic vision, operational efficiency, and innovation in AI infrastructure and cloud services.
Executive Leadership
- Michael Intrator: Co-Founder and Chief Executive Officer
Senior Executives
- Brannin McBee: Founder & Chief Development Officer
- Brian V.: Chief Strategy Officer
- Chen Goldberg: Senior Vice President of Engineering (former General Manager & VP of Engineering for Kubernetes and Serverless at Google)
- Sachin Jain: Chief Operating Officer (previously Senior Vice President at Oracle Cloud, with roles at Google Cloud and Amazon)
- Chetan Kapoor: Chief Product Officer (former member of the Compute leadership team at Amazon Web Services)
- Nitin Agrawal: Chief Financial Officer (former Vice President of Finance at Google Cloud)
- Jeff Baker: Chief Accounting Officer (played a crucial role in Rivian's IPO)
Other Key Roles
- Marco Iodice: Co-Founder
- Nicholas Hoover: Vice President, Corporate Development
- Michelle O'Rourke: Chief People Officer (promoted from Senior Director of Human Resources) This diverse team of leaders brings extensive experience from top tech companies, positioning CoreWeave for continued growth and innovation in AI infrastructure and cloud services.
History
CoreWeave's journey from a cryptocurrency mining operation to a leading cloud computing provider specializing in GPU infrastructure for AI and ML workloads is a testament to its adaptability and strategic vision.
Founding and Early Days (2017-2018)
- Founded in 2017 by Michael Intrator, Brian Venturo, and Brannin McBee
- Initially known as Atlantic Crypto, focusing on Ethereum mining using GPUs
- Founders were previously commodities traders at Hudson Ridge Asset Management LLC
Pivot and Transition (2018-2019)
- Shifted focus from cryptocurrency mining following the 2018 market crash
- Raised $1.2 million seed round in 2019
- Diversified into cloud infrastructure services
- Acquired Leonardo Render and launched Concierge Render, marking entry into cloud infrastructure business
Expansion and Growth (2019-2024)
- Rebranded from Atlantic Crypto to CoreWeave in October 2021
- Specialized in providing cloud-based GPU infrastructure
- Developed Kubernetes-native architecture for large-scale, GPU-intensive tasks
- Partnered with NVIDIA to offer high-performance computing solutions
Funding and Valuation Milestones
- August 2023: Secured $2.3 billion debt financing facility
- May 2024: Raised $1.1 billion, reaching a $19 billion valuation
- October 2024: Valuation increased to $23 billion
- November 2024: Closed a $650 million secondary share sale
Expansion Plans and Partnerships
- Announced construction of a $1.2 billion data center in New Jersey
- Planned £750 million investment in the U.K.
- Multi-year deal with Microsoft worth billions of dollars
- Contract with Core Scientific for additional infrastructure
Current Status (2024)
- Operates 13 data centers in the U.S. and 2 in the U.K.
- Employs approximately 550 people
- Led by Michael Intrator (CEO), Brian Venturo (CTO), and Brannin McBee (CSO)
- Planning IPO for 2025 CoreWeave's history demonstrates its ability to adapt to market changes and capitalize on emerging opportunities in the AI and cloud computing sectors.
Products & Solutions
CoreWeave, a specialized cloud provider, offers a range of high-performance computing solutions tailored for sectors such as machine learning, artificial intelligence, and visual effects. Their key products and solutions include:
GPU and CPU Compute Solutions
CoreWeave provides GPU-accelerated and CPU compute resources essential for compute-intensive tasks. They are among the first to market with the latest NVIDIA GPUs at supercomputing scale, enabling faster deployment of AI workloads.
Managed Kubernetes
The company offers managed Kubernetes services for efficient deployment and management of containerized applications, particularly useful for AI and ML workloads requiring scalable infrastructure.
Virtual Servers and Storage
CoreWeave's virtual servers and cloud storage solutions are designed to handle complex AI workloads. Their storage services feature strategic checkpointing, automated snapshots, and high reliability to ensure data integrity and minimal downtime.
Networking
High-speed networking services are provided to meet the high-bandwidth requirements of AI and ML applications, ensuring a secure environment for mission-critical AI development.
VFX & Rendering Solutions
In collaboration with Conductor, CoreWeave offers flexible on-demand artist workstations and virtually unlimited rendering capacity for visual effects, animation, motion graphics, and design projects. This modernizes VFX workflows by reducing infrastructure management and render wait times.
Security and Connectivity
The cloud platform is built with enterprise-grade security features, including encryption at rest and in transit, identity access management protocols, and role-based policies to ensure data protection.
Pixel Streaming
CoreWeave supports pixel streaming, beneficial for real-time rendering and other graphics-intensive applications, allowing for efficient and scalable computing resources. CoreWeave's products and solutions are optimized for AI workloads, providing a resilient cloud platform that accelerates time-to-market and enhances the efficiency of compute-intensive tasks across various industries.
Core Technology
CoreWeave's core technology and infrastructure are specialized for high-performance, GPU-intensive workloads, particularly in artificial intelligence (AI). Key aspects of their technology include:
GPU Infrastructure
CoreWeave provides cloud-based GPU infrastructure with a large inventory of GPUs, including the latest NVIDIA H100 GPUs. This infrastructure is designed to support large-scale, GPU-accelerated workloads.
Kubernetes-Native Architecture
The company's infrastructure is built on a Kubernetes-native architecture, enabling efficient management and scaling of resources to support the complex requirements of AI and machine learning workloads.
Customized Solutions
Unlike traditional cloud providers, CoreWeave's infrastructure is hyper-optimized for AI workloads. Their cloud services are built from the ground up to handle the massive parallel processing power and memory requirements of AI applications, offering superior efficiency and performance.
Automation and Support
CoreWeave combines automated recovery processes with expert engineering support to minimize downtime and ensure swift resolution of failures, guaranteeing high availability for critical AI applications.
Enterprise-Grade Security and Connectivity
The company offers robust security capabilities and high-speed connectivity, providing a secure and reliable environment for building mission-critical AI applications.
Data Centers and Expansion
As of 2024, CoreWeave operates thirteen data centers across the United States and two in the United Kingdom, with plans for further expansion. This includes a significant investment in the U.K. and the construction of a new data center in New Jersey.
Partnerships and Collaborations
CoreWeave has significant partnerships, including a major contract with Core Scientific for high-performance computing (HPC) hosting. This partnership involves modifying existing infrastructure to support dense HPC operations, highlighting CoreWeave's commitment to providing specialized infrastructure for AI and HPC workloads.
Industry Peers
CoreWeave, a specialized cloud provider focused on GPU-accelerated compute resources, operates in a competitive landscape with several industry peers across various segments:
Cloud and Hyperscale Providers
- DigitalOcean: A cloud provider used as a comparison for CoreWeave's IPO valuation
- Amazon Web Services (AWS), Google Cloud Platform (GCP), and Azure: Major hyperscalers serving as benchmarks for CoreWeave's market positioning
AI and Machine Learning Specialists
- Graphcore: Specializes in processor technology for AI and machine learning
- Tenstorrent: Competes in the AI and machine learning space
- Lambda: Provides computation solutions for deep learning and AI processes
GPU and High-Performance Computing (HPC)
- NVIDIA: A key partner for CoreWeave, sometimes used as a valuation comparison point
- Arc Compute: Specializes in GPU resource optimization
Other Competitors
- Soluna Cloud, Foundry, DataCrunch, and Radium: Offer various cloud computing and HPC solutions
- Lumino: Provides cloud-based solutions for AI model development and training
Data Center and Infrastructure Companies
- Arista Networks, Equinix, and NextDC: Involved in data center operations and infrastructure These companies span different segments but are all relevant to CoreWeave's specialized focus on GPU-accelerated cloud computing and high-performance computing resources. The competitive landscape highlights the growing demand for specialized AI and HPC infrastructure, with CoreWeave positioning itself as a leader in providing tailored solutions for compute-intensive applications.