Blog

CoreWeave acquires Core Scientific to boost operational efficiency and power the future of high-performance computing and AI innovation

CoreWeave launches the NVIDIA GB300 NVL72, delivering record-breaking AI inference performance and setting a new standard for next-gen, large-scale AI infrastructure.

Discover how CAIOS achieves 2+ GB/s per GPU throughput—redefining performance for AI model training at scale.

CoreWeave's MLPerf v5.0 results set new AI performance records, helping customers innovate faster with optimized, cost-efficient infrastructure.

Learn how Jane Street gained a flexible, full-stack solution that would meet its demands for scalability, efficiency, and security through CoreWeave.

This collaboration established a new benchmark for speed and scalability in foundational model training.

Today we announce the expansion of our NVIDIA Blackwell-based instances with the general availability of NVIDIA HGX B200-based instances on CoreWeave

Insights from SemiAnalysis’s ClusterMAX™ Report and how CoreWeave Set the Bar for AI Cloud Excellence

We're proud to announce we've joined Red Hat's new llm-d OSS project as a founding contributor. Learn more about how it's transforming AI inference.

AI leadership starts with infrastructure. CoreWeave CEO Mike Intrator tells the Senate why building AI at home is key to U.S. competitiveness.

CoreWeave has acquired Weights & Biases to deliver tighter AI workflow integration—combining infrastructure and experiment tracking for faster, more efficient model development at scale.

Discover how CoreWeave enables auto-scaling of Deadline workflows and tasks on-demand, providing VFX and animation studios with flexible, cost-effective cloud resources.

Learn how to launch a GPU-enabled Windows cloud desktop with Parsec, enabling low-latency remote access to high-performance compute resources for creative, AI, and engineering applications.

Learn how Determined AI and CoreWeave Cloud enable faster, more efficient training of deep learning models, offering scalable infrastructure, easy experiment management, and optimized performance.

CoreWeave partners with Spire Animation Studios to enable collaborative filmmaking, providing scalable cloud resources that streamline production workflows and accelerate creative collaboration.

Learn how to quickly create and launch a Jupyter Notebook on CoreWeave Cloud, enabling developers and researchers to build, test, and scale AI and machine learning projects.

Learn how to get started with platform engineering, offering insights into building scalable, efficient cloud platforms that simplify deployment and infrastructure management.

NovelAI CEO Eren Dogan joins CoreWeave for a webinar discussing how NovelAI leverages CoreWeave’s infrastructure to accelerate AI model training, optimize performance, and scale innovations.

CoreWeave executives break down the company's key value propositions, highlighting its leadership in AI infrastructure, unmatched performance, scalability, and support for AI-driven enterprises.

CoreWeave offers the performance, scalability, and specialized infrastructure needed to host AI and machine learning projects, delivering faster training, lower costs, and expert support.

Perception leverages Conductor to deliver high-quality visual effects with faster rendering times, enhanced scalability, and the flexibility needed to meet the demands of complex creative projects.

Discover how Conductor delivers faster, more scalable cloud rendering for VFX and animation, offering artists and studios greater flexibility, lower costs, and seamless creative workflows.

CoreWeave’s VFX showreel features stunning animations from films like Deadpool, Shang-Chi, and Lightyear, with all shots rendered using Conductor’s cloud-based platform for artists and studios.

Nick Ihli from SchedMD discusses the power of Slurm workload manager at Supercomputing, highlighting its role in efficiently orchestrating large-scale AI and HPC clusters across diverse environments.

Explore why bare metal infrastructure provides a superior foundation for serverless Kubernetes, delivering better performance, lower latency, and greater control for demanding AI and cloud workloads.

Zeet and CoreWeave teamed up at Supercomputing to showcase how their collaboration simplifies deployment pipelines, enabling faster, more efficient scaling of AI and cloud-native applications.

At Supercomputing 2022, CoreWeave highlighted how Vast Data’s high-performance storage solutions support AI workloads by enabling greater scalability, efficiency, and speed.

Andy Pernsteiner of CoreWeave presents at Supercomputing 2023, showcasing how Vast Data solutions enhance AI infrastructure with scalable, high-performance storage.

CoreWeave CBO Mike Mattacola joins NVIDIA’s David Hogan and The Times’ Katie Prescott to discuss AI growth in Europe.

CoreWeave CEO Mike Intrator joins Bloomberg’s Tom McKenzie for a Fireside Chat on AI at London Tech Week.

Get a glimpse into life at CoreWeave and discover our inclusive, supportive, and exciting company culture.

Chief Strategy Officer Brian Venturo shares insights on the future of data centers and CoreWeave’s fast-growing data center footprint.

Check out one of the first GB200 NVL72 demos, highlighting its performance, cooling tech, networking strength, and energy efficiency.

CoreWeave, the AI Hyperscaler™, offers a cloud platform with advanced software to fuel the next wave of AI, delivering accelerated computing solutions for enterprises and top AI labs.

Watch how CoreWeave drives global AI innovation with a cloud platform built for compute-intensive workloads.

CoreWeave Mission Control handles cluster health management for you, delivering top-tier reliability and resiliency with a typical node goodput of 96% for AI infrastructure.

Liquid cooling: Discover why liquid cooling is critical for AI at scale—unlocking higher GPU density, energy efficiency, and sustained performance for the most demanding workloads.

See how CoreWeave stays the top AI cloud platform, designed for the performance, scale, and expertise needed to fuel AI innovation and support accelerated computing demands.

Learn about what object storage needs for LLMs: fast access, quick recovery, scalability, and strong security—and how CAIOS provides it all.

CoreWeave is the sole AI cloud provider to earn SemiAnalysis’s top-tier Platinum ClusterMAX™ rating, recognized for superior performance, reliability, and scalability in large-scale GPU clusters.

Trillion Labs exponentially scales its AI models with CoreWeave’s H100 clusters, gaining faster time-to-market, cost efficiency, and resilient infrastructure support.

CoreWeave achieved top MLPerf v5.0 AI inference results, delivering 800 TPS on Llama 3.1 405B with NVIDIA GB200 and 33,000 TPS on Llama 2 70B with H200 GPUs, marking significant performance gains.

Learn more about 3 strategies to boost AI infrastructure goodput to 96%: using optimized hardware, proactive interruption management, and efficient checkpointing for fast recovery.

Check out our GTC 2025 recap, highlighting AI breakthroughs, new partnerships, keynotes, tech demos, and major product launches like CoreWeave AI Object Storage and NVIDIA GB200 NVL72 racks.

Learn more about the launch of CoreWeave AI Object Storage, a high-performance, scalable, and secure storage service built to accelerate AI training and inference with seamless GPU integration.

CoreWeave boosts AI infrastructure efficiency with up to 20% higher GPU cluster performance compared to alternatives, narrowing the AI efficiency gap with optimized, purpose-built solutions.

CoreWeave will soon launch instances with NVIDIA RTX PRO 6000 Blackwell GPUs, offering major AI and graphics performance boosts, enhanced efficiency, and seamless data access.

CoreWeave now supports NVIDIA AI Enterprise and Cloud Functions, helping customers easily deploy, scale, and optimize AI applications on high-performance cloud infrastructure.

CoreWeave is acquiring Weights & Biases to create an end-to-end AI platform, combining compute and MLOps tools to speed AI model development and deployment.

CoreWeave unveils a GenAI-optimized networking stack featuring NVIDIA Quantum-2 InfiniBand, SHARP, and SHIELD technologies, delivering ultra-low latency, high throughput, and scalable GPU clusters.

Learn about six key strategies that optimize high-performance computing storage for machine learning, boosting data access speed, pipeline efficiency, and overall AI training performance.

CoreWeave will co-sponsor the Inference-Time Compute AI Hackathon in San Francisco, supporting teams competing to build next-gen AI reasoning apps with up to $60,000 in prizes.

Our AI-optimized data centers feature purpose-built infrastructure, including NVIDIA Quantum-2 InfiniBand networking and liquid cooling for high-density, low-latency AI workloads.

Learn more about 4 strategies for full-stack AI observability that boost infrastructure reliability, optimize performance, prevent failures, and speed up AI model development.

Discover how distributed file storage enhances AI model training by delivering the performance, scalability, and reliability needed to handle large datasets and accelerate development cycles.

CoreWeave is the first cloud provider to offer NVIDIA GB200 NVL72 instances, delivering 1.44 exaFLOPS of AI compute and 13.5TB of NVLink-connected memory for next-gen AI workloads.

CoreWeave is the first cloud provider to offer NVIDIA GB200 NVL72 instances, delivering up to 1.44 exaFLOPS of AI compute and 13.5TB of NVLink-connected memory per rack.

Learn how we're expanding our AI data centers with liquid cooling, high-density racks, and scalable power to meet surging AI compute demands efficiently.

Learn how CoreWeave is the AI hyperscaler—setting the standard with cutting-edge GPUs, reliable infrastructure, world-class observability, and elite managed services.

We showcased NVIDIA GB200 NVL72 AI performance, energy efficiency, and cooling innovation, backed by Kubernetes support, observability tools, and Quantum-2 InfiniBand networking.

Check out our recap of KubeCon 2024, where we showcased how to manage Kubernetes-based AI clusters, demos of Mission Control, and insights into scaling AI workloads.

See how SUNK layers Slurm and Kubernetes environments to boost GPU utilization, simplify scaling, and speed AI development by balancing training and inference workloads on a single cluster.

The 2024 LLM market saw explosive growth in model size, efficiency, and multimodality, with open-source innovation and massive AI infrastructure scaling driving the next wave of advancements.

CoreWeave advances AI infrastructure with one of the first NVIDIA GB200 NVL72 clusters, new GH200, L40, L40S GPU instances, and previews AI Object Storage to boost AI and HPC workloads.

See how OpenAI's Strawberry model (o1) advances LLM reasoning, solving complex tasks with high accuracy—83.3% in programming and 100% in math—marking a step toward AGI.

Learn how CoreWeave Mission Control boosts cluster management with continuous monitoring, transparency, and proactive reliability to speed AI model development and minimize downtime.

CoreWeave and Run:ai have partnered to integrate Run:ai’s orchestration platform with CoreWeave’s high-performance AI infrastructure.

CoreWeave is the first cloud provider to deploy NVIDIA H200 Tensor Core GPUs, offering industry-leading AI performance with 4.8TB/s bandwidth and 141GB HBM3e memory for GenAI workloads.

See how our strategic hiring of of industry veterans has been pivotal in scaling our AI cloud infrastructure.

CoreWeave is among the first to deploy NVIDIA Blackwell clusters, offering massive AI performance gains with HGX B200 and GB200 NVL72 systems.

Check out 4 MLOps best practices for AI clusters: regular node health checks, continuous monitoring, fast data loading, and better orchestration to boost reliability, speed, and efficiency.

Conductor’s cloud-agnostic rendering lets studios scale across providers, access more compute options, and boost efficiency.

CoreWeave partnered with Authentik to streamline onboarding and simplify identity management for developers at a major AI/ML hackathon, accelerating participation and project success.

CoreWeave was honored as one of USA Today's Top Workplaces for 2024, recognizing the company's commitment to fostering a strong, supportive culture across all areas of its growing organization.

As AI demand accelerates, data centers are being redesigned to prioritize greater power, cooling efficiency, density, and resiliency, supporting the needs of AI and accelerated computing environments.

Mistral AI and CoreWeave highlighted their collaboration at NVIDIA GTC, demonstrating how their partnership advances cutting-edge AI model development and innovation for global enterprises.

UbiOps and CoreWeave partnered to combine high-performance compute with a powerful deployment platform, helping companies efficiently productize and scale AI models for real-world applications.

A new native Blender plug-in for Conductor streamlines workflows for artists and creators, making it easier to manage rendering tasks, improve efficiency, and accelerate creative projects.

CoreWeave and Loft Labs use vCluster technology to run virtual Kubernetes clusters at scale, improving resource efficiency, speeding up deployment times, and enhancing flexibility for AI workloads.

Decart and Cerebrium announced their commitment to empowering the next million users with large language model applications, making AI more accessible, scalable, and impactful.

CoreWeave kicked off SC23 by announcing plans to offer new NVIDIA GH200 Grace Hopper Superchip-powered instances in early 2024, bringing enhanced performance and efficiency to AI and HPC workloads.

CoreWeave's SUNK, a Slurm on Kubernetes implementation, simplifies running HPC and large-scale AI workloads by combining the orchestration power of Kubernetes with Slurm’s scheduling efficiency.

Learn how autoscaling impacts compute costs for inference, enabling AI applications to dynamically adjust resources, optimize performance, and reduce expenses.

Explore why specialized infrastructure is critical for powering AI, as organizations move beyond traditional compute to support the energy, scalability, and performance demands of AI workloads.

CoreWeave collaborated with Product Insight to deliver GPU-accelerated medical animations, enabling faster rendering, greater visual fidelity, and enhanced storytelling.

Explore Conductor's rendering benchmarks for GPU and CPU instances, showcasing significant performance gains, cost efficiencies, and scalability advantages for VFX and animation workflows.

CoreWeave’s new Datadog integration enables users to monitor, visualize, and optimize their cloud usage in real time, improving operational efficiency, cost management, and performance.

Learn how NovelAI trained Clio, a highly performant 3-billion-parameter NLP model, on CoreWeave’s cloud infrastructure, showcasing efficiency, scalability, and speed.

CoreWeave secured a $2.3 billion debt financing facility led by Magnetar Capital and Blackstone, fueling the company’s rapid expansion and strengthening its position as a leader.

Learn how to choose the best GPU instances for rendering on Conductor, with insights on balancing performance, cost, and scalability to optimize visual effects, animation, and creative workflows.

Odyssey powered McDonald’s first metaverse experience using CoreWeave, delivering the scalability, performance, and reliability needed to support an immersive digital environment.

CoreWeave and NVIDIA achieved record-breaking MLPerf results with their cloud-native AI supercomputer, demonstrating unmatched performance for training and inference workloads.

CoreWeave’s Tensorizer accelerates PyTorch model load times by optimizing memory management and streamlining deployment, helping developers reduce startup latency and improve performance.

Discover how Eleuther AI leveraged NVIDIA Triton Inference Server and CoreWeave’s infrastructure to efficiently serve LLM inference at scale, optimizing performance, speed, and resource utilization.