{"slug": "subaru-reduced-30gb-container-pulls-from-three-hours-to-three-minutes", "title": "Subaru reduced 30GB container pulls from three hours to three minutes", "summary": "Subaru reduced pull times for 30+ GB AI container images from approximately three hours to three minutes, a 60x improvement, by optimizing its Kubernetes networking architecture using Envoy Gateway. The Cloud Native Computing Foundation announced Subaru as the winner of the CNCF End User Case Study Contest at KubeCon + CloudNativeCon Japan 2026 for transforming the AI infrastructure behind its next-generation EyeSight development platform. By adopting GitOps with Argo CD and Helmfile, and automating ML pipelines with Argo Workflows, Subaru standardized deployments and improved operational efficiency for its advanced driver assistance systems.", "body_md": "*New architecture reduced AI container image pull times by 60x while automating workflows for next-generation driver assistance systems*\n\n**Key Highlights**\n\n- Subaru reduced pull times for 30+ GB AI container images from approximately three hours to three minutes, a 60x improvement, by optimizing its Kubernetes networking architecture using Envoy Gateway.\n- By adopting GitOps with Argo CD and Helmfile, Subaru standardized application deployments and improved reproducibility across its AI development environments.\n- Using Argo Workflows, Subaru automated its ML pipelines, improving operational efficiency while accelerating development of next-generation EyeSight, its advanced driver assistance systems (ADAS).\n\n**YOKOHAMA – KUBECON + CLOUDNATIVECON JAPAN — July 29, 2026 — **[The Cloud Native Computing Foundation® ](https://www.cncf.io/)(CNCF®), which builds sustainable ecosystems for cloud native software, today announced [Subaru](https://www.subaru.co.jp/en/), one of the world’s leading automobile manufacturers, as the winner of the CNCF End User Case Study Contest for [KubeCon + CloudNativeCon Japan 2026](https://events.linuxfoundation.org/kubecon-cloudnativecon-japan/).\n\nSubaru received the award for transforming the AI infrastructure behind its next-generation EyeSight development platform. By combining [Kubernetes](https://www.cncf.io/projects/kubernetes/) with multiple CNCF technologies such as [Envoy Gateway](https://www.cncf.io/blog/2022/05/16/introducing-envoy-gateway/), [Gateway API](https://gateway-api.sigs.k8s.io/docs/introduction/) and [MetalLB](https://www.cncf.io/projects/metallb/), the company dramatically improved developer productivity, infrastructure efficiency and machine learning (ML) reproducibility. By optimizing the delivery of large AI container images, adopting GitOps practices and automating end-to-end ML workflows, Subaru created a scalable cloud native platform that enables faster iteration on AI models while improving operational consistency.\n\n“As organizations build increasingly sophisticated AI applications, cloud native technologies provide the foundation for making those workloads scalable and operationally efficient,” said Chris Aniszczyk, CTO, CNCF. “Subaru demonstrates how combining Kubernetes with projects like Argo, Envoy, Helm, Harbor and MetalLB can solve real world infrastructure challenges while accelerating AI innovation. It’s an excellent example of open source cloud native technologies delivering measurable business impact.”\n\nDeveloping AI models for Subaru’s next-generation EyeSight required continuous cycles of data processing, model training, validation and inference. As those workloads grew in scale and complexity, Subaru’s existing on-premises GPU environment became a bottleneck. AI container images exceeding 30 GB routinely required hours to download before workloads could begin. Deployment processes relied on manually executed scripts rather than declarative configuration and complex machine learning pipelines lacked a unified orchestration framework. All made it difficult to achieve reproducible, efficient execution.\n\n“We wanted to build an AI development platform that removed operational friction and enabled our engineers to focus on improving model accuracy instead of managing infrastructure,” said Ryoji Kobayashi, DevOps engineer of ADAS development department, Subaru. “By embracing cloud native technologies and GitOps practices, we’ve significantly reduced development bottlenecks while creating a more reproducible, scalable platform for machine learning. The improvements we’ve achieved are helping us accelerate innovation for next-generation EyeSight.”\n\nTo address these challenges, Subaru built a cloud native AI platform on Kubernetes using multiple CNCF projects. By combining [Envoy Gateway](https://www.cncf.io/blog/2022/05/16/introducing-envoy-gateway/), [Gateway API](https://gateway-api.sigs.k8s.io/docs/introduction/) and [MetalLB](https://www.cncf.io/projects/metallb/), the company significantly reduced pull times for large AI container images from approximately three hours to just three minutes; an impressive 60x improvement. Subaru also adopted a GitOps workflow with [Argo CD](https://www.cncf.io/projects/argo/) and Helmfile, building on its existing [Helm](https://www.cncf.io/projects/helm/)-based deployments to standardize application delivery and centrally manage application definitions. Argo Workflows automated the end-to-end machine learning pipeline, improving reproducibility and operational efficiency.\n\nSubaru continues to expand its use of Kubernetes and CNCF technologies to advance its AI infrastructure. As AI workloads become more complex, the company plans to build on its cloud native foundation to further improve developer productivity, platform capabilities and operational efficiency.\n\nSubaru’s achievement was announced on stage during a keynote session at KubeCon + CloudNativeCon Japan, where Ryoji Kobayashi delivered a [keynote](https://events.linuxfoundation.org/kubecon-cloudnativecon-japan/program/schedule/?id=1283173) to showcase the details of the winning use case.\n\nTo review the complete architectural implementation and downstream results, read the full [Subaru case study](https://www.cncf.io/case-studies/subaru/).\n\n**Additional Resources**\n\n**About Cloud Native Computing Foundation**\n\nCloud native computing empowers organizations to build and run scalable applications with an open source software stack in public, private, and hybrid clouds. The Cloud Native Computing Foundation (CNCF) hosts critical components of the global technology infrastructure, including Kubernetes, Prometheus, and Envoy. CNCF brings together the industry’s top developers, end users, and vendors and runs the largest open source developer conferences in the world. Supported by nearly 800 members, including the world’s largest cloud computing and software companies, as well as over 200 innovative startups, CNCF is part of the nonprofit Linux Foundation. For more information, please visit [www.cncf.io](http://www.cncf.io).\n\n**Media Contact**", "url": "https://wpnews.pro/news/subaru-reduced-30gb-container-pulls-from-three-hours-to-three-minutes", "canonical_source": "https://www.cncf.io/announcements/2026/07/28/subaru-wins-cncf-end-user-case-study-contest-for-accelerating-ai-development-with-cloud-native-infrastructure/", "published_at": "2026-07-29 08:59:31+00:00", "updated_at": "2026-07-29 09:22:23.045103+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "ai-infrastructure", "developer-tools", "ai-research"], "entities": ["Subaru", "Cloud Native Computing Foundation", "Envoy Gateway", "Kubernetes", "Argo CD", "Helmfile", "Argo Workflows", "EyeSight"], "alternates": {"html": "https://wpnews.pro/news/subaru-reduced-30gb-container-pulls-from-three-hours-to-three-minutes", "markdown": "https://wpnews.pro/news/subaru-reduced-30gb-container-pulls-from-three-hours-to-three-minutes.md", "text": "https://wpnews.pro/news/subaru-reduced-30gb-container-pulls-from-three-hours-to-three-minutes.txt", "jsonld": "https://wpnews.pro/news/subaru-reduced-30gb-container-pulls-from-three-hours-to-three-minutes.jsonld"}}