Creative · Multi-Cloud
Adobe
Adobe runs a **multi-cloud** architecture primarily across AWS, Azure, and selected GCP sovereign regions, with Adobe Experience Platform and Experience Cloud services deployed on both AWS and Azure, and Firefly’s GPU-heavy training and inference infrastructure built on AWS. Core data and AI workloads rely on managed storage, compute, and networking primitives from these clouds, with Adobe layering proprietary platforms such as Experience Platform, Sensei, and Firefly on top.[1][2][3][5][6][7][12][14][15]
Foundation proximity: Native (own / self-hosted weights)
Foundational models & how they consume them
PrimaryAdobe Firefly (family of diffusion and generative models, including Image 3 and Firefly Foundry variants)· Adobe (self-hosted on AWS with NVIDIA GPU infrastructure; Adobe Sensei/Firefly platform) · Native (own / self-hosted weights)
SecondaryCustom/bespoke Firefly Foundry models (brand-tuned generative models)· Adobe (co-developed with customers, deployed within Adobe ecosystem) · Native (own / self-hosted weights)
TertiaryNVIDIA-based Firefly stack (CUDA-X, NeMo, Cosmos stack for Firefly models)· NVIDIA (infrastructure and model/tooling stack underpinning Firefly training and deployment) · Native (own / self-hosted weights)
Cloud & datastore
Multi-CloudAmazon S3 (data lake/object storage for AI training and inference)[14]Amazon FSx for Lustre (high-performance file storage for GPU training workflows)[14]Managed storage and data services on AWS (EBS, EFA, etc., used as part of AI training platform)[14]Azure-based managed data stores and PaaS services for Adobe Experience Platform and Commerce (multi-cloud deployment on Azure and AWS)[1][2][7][12]GCP sovereign regions for Adobe Experience Manager Managed Services (data residency-focused deployments)[5]Proprietary Adobe data services layered on top of these cloud providers (Experience Platform and Experience Cloud foundations)[6][15]Compliance
Digital sovereignty controls leveraging AWS, Azure, and GCP sovereign regions for Adobe Experience Manager Managed Services (jurisdictional control, data residency, added encryption and controls)[5]Multi-region, high-availability architecture across AWS or Azure Availability Zones for Adobe Commerce on cloud (disaster recovery and resilience)[7]Content provenance and authenticity via Content Credentials using C2PA standards for AI-generated content (Firefly-related trust and safety)[4]Others that build on proprietary / self-built models
Abnormal SecurityAbridgeAI21 LabsAleph AlphaAmbienceAT&TExplore
FAQ
Does Adobe use a foundational AI model?
Adobe uses Adobe Firefly (family of diffusion and generative models, including Image 3 and Firefly Foundry variants), Custom/bespoke Firefly Foundry models (brand-tuned generative models), NVIDIA-based Firefly stack (CUDA-X, NeMo, Cosmos stack for Firefly models) (Native (own / self-hosted weights)).
What cloud does Adobe run on?
Adobe runs on Multi-Cloud.
What database does Adobe use?
Adobe uses Amazon S3 (data lake/object storage for AI training and inference)[14], Amazon FSx for Lustre (high-performance file storage for GPU training workflows)[14], Managed storage and data services on AWS (EBS, EFA, etc., used as part of AI training platform)[14], Azure-based managed data stores and PaaS services for Adobe Experience Platform and Commerce (multi-cloud deployment on Azure and AWS)[1][2][7][12], GCP sovereign regions for Adobe Experience Manager Managed Services (data residency-focused deployments)[5], Proprietary Adobe data services layered on top of these cloud providers (Experience Platform and Experience Cloud foundations)[6][15].
How close does Adobe run to the foundational model?
Adobe is Native (own / self-hosted weights) — close to the metal, so cheaper tokens and more control over outputs.
Sources
adobe.com ↗Also on the map
← Full orbit map (260 companies) · Score your own stack →
Architecture inferred from public sources · confidence medium · verify before betting on a detail.