RUN:AI
Tel Aviv, Israel
$700M
VALUATION
techcrunch.comAbout
Israeli GPU orchestration platform for ML workloads, acquired by Nvidia. Optimizes GPU utilization and resource allocation for AI training.
About Run:ai
Israeli GPU orchestration platform for ML workloads, acquired by Nvidia. Optimizes GPU utilization and resource allocation for AI training.
AI Focus Areas
- GPU virtualization and scheduling
- AI workload orchestration across clouds
- Cluster utilization optimization
- Multi-tenant AI infrastructure management
- Enterprise AI platform integration
Key Products
- Run:ai Atlas platform
- Kubernetes-based GPU scheduler
- Utilization and capacity analytics
- Integrations into Nvidia DGX and DGX Cloud
Market Position
Run:ai carved out a niche by making GPU resources fungible across on-prem, cloud and edge clusters, abstracting hardware details from data science teams. By virtualizing GPUs and intelligently scheduling jobs, it helped enterprises significantly increase utilization, a strong ROI story given expensive accelerators. Pre-acquisition, it competed with internal tooling, cloud-native schedulers and a small set of specialized vendors. Nvidia’s purchase validates its importance and folds its capabilities into DGX Cloud, giving Run:ai privileged access to Nvidia’s customer base and roadmap. As part of Nvidia, it may shift from vendor-neutral to Nvidia-centric, but it gains scale and distribution as AI workloads explode.
AGI Relevance
At AGI scales, efficient use of compute becomes as important as raw capacity. Systems like Run:ai’s scheduler determine how quickly models can be trained, how experiments are parallelized and how expensive infrastructure is amortized. Fine-grained GPU sharing and dynamic allocation are key for running large numbers of concurrent experiments and agentic workloads. By optimizing these layers, Run:ai effectively increases the ‘effective FLOPs’ available to research labs and enterprises. Its integration into Nvidia’s stack means that improvements can be propagated widely, influencing the cost and feasibility of AGI-class training and inference for many organizations.
Investment Highlights
Raised a $30M Series B in 2021 led by Insight Partners (following a $75M Series C in 2022 per TechCrunch) and a total of over $100M in venture funding. In April 2024, Nvidia agreed to acquire Run:ai for around $700M, according to TechCrunch, marking a strong exit in the AI infrastructure space.
Tags
- GPU Management
- MLOps
- Kubernetes
Recent News
No recent news for Run:ai