CoreWeave Launches Forge Platform for Open Full-Stack AI Cloud

CoreWeave Inc. has announced the launch of CoreWeave Forge, a new development platform designed to connect AI training, inference, and evaluation processes. According to a report by SiliconANGLE, the company aims to use insights gained from production environments to continuously improve artificial intelligence models and agents.

Jean English, chief marketing officer at CoreWeave, stated that the development loop should remain open to different models, frameworks, and cloud environments. She emphasized that the platform is not closed, allowing teams to leverage what they learn in production to enhance their AI systems. This approach seeks to distinguish CoreWeave in the full-stack AI cloud market.

Building Infrastructure for AI Workloads

CoreWeave built its platform specifically for AI workloads rather than adapting infrastructure designed for earlier generations of web applications. English explained that clients migrating from other cloud providers often encounter gaps in speed, performance, capacity, or tooling. The company argues that the AI era requires building from first principles to achieve the necessary performance levels.

The infrastructure was constructed from the ground up to address these specific needs. English noted that clients using other clouds may struggle to move as fast or achieve the performance required for modern AI tasks. This purpose-built design is central to CoreWeave’s strategy in the competitive cloud market.

Hardware Validation and Partner Ecosystems

Hardware remains a critical component of CoreWeave’s offering, with the company being the first to bring up and validate Nvidia Corp.’s Vera Rubin platform. Additionally, CoreWeave recently earned SemiAnalysis’ Platinum ClusterMAX rating for the third consecutive time. These achievements highlight the company’s focus on high-performance computing capabilities for its customers.

However, English indicated that the opportunity extends beyond the graphics processing unit itself. The full-stack AI cloud concept includes tooling, infrastructure, and application programming interface connections supplied by various partners. These partners enable access to a wide range of models, supporting the open development loop that CoreWeave promotes.

What should you do?

Security and technology professionals should evaluate how new AI development platforms integrate with existing security frameworks. Organizations adopting full-stack AI clouds need to verify that vendor tooling supports their specific compliance and data governance requirements.

WATCH FOR:

• Vendor announcements regarding new AI infrastructure features and partner integrations.
• Updates on hardware validation statuses for new GPU platforms.
• Changes in cloud provider strategies regarding open versus closed AI ecosystems.
• Industry ratings and benchmarks for AI cluster performance and efficiency.

What to do and how to stay safe: CoreWeave

• Review current AI infrastructure contracts to ensure they support open model frameworks.
• Verify that development platforms allow integration with multiple cloud environments.
• Monitor vendor roadmaps for updates on hardware validation and partner tooling.
• Assess whether existing cloud providers meet specific AI workload performance needs.

Step-by-step guide: Cloud Misconfiguration Response: Containment, Recovery and Prevention Steps

General security guidance from the Threat Times newsroom. It is not confirmed advice from the organisations named in this story.

Frequently asked questions

What is CoreWeave Forge?

CoreWeave Forge is a newly announced development platform that connects AI training, inference, and evaluation to improve models and agents in production.

Why does CoreWeave emphasize an open loop?

The company believes the development loop should be open to different models, frameworks, and clouds, rather than being closed to a single ecosystem.

What hardware achievements has CoreWeave reported?

CoreWeave was the first to validate Nvidia’s Vera Rubin platform and recently earned SemiAnalysis’ Platinum ClusterMAX rating for a third consecutive time.