Integrated supercomputing architecture
AI Hypercomputer
AI optimized hardware, software, and consumption, combined to improve productivity and efficiency.
Overview
Performance-optimized hardware
Our performance optimized infrastructure including Google Cloud TPU , Google Cloud GPU , Google Cloud Storage and the underlying Jupiter network consistently provides fastest time to train for large scale state of the art models due to the strong scaling characteristics of the architecture which leads to the best price/performance for serving large models.
Open software
Our architecture is optimized to support the most common tools and libraries such as Tensorflow, Pytorch and JAX . Plus it allows customers to take advantage of technologies such as Cloud TPU Multislice and Multihost configurations and managed services like Google Kubernetes Engine . This allows customers to deliver turnkey deployment for common workloads like the NVIDIA NeMO framework orchestrated by SLURM .
Flexible consumption
Our flexible consumption models allow customers to choose fixed costs with committed use discounts or dynamic on-demand models to meet their business needs. Dynamic Workload Scheduler helps customers get the capacity they need without over allocating so they are only paying for what they need. Plus, Google Cloud's cost optimization tools help automate resource utilization to reduce manual tasks for engineers.
How It Works
Google is a leader in artificial intelligence with the invention of technologies like TensorFlow. Did you know you can leverage Google’s technology for your own projects? Learn about Google's history of innovation in AI infrastructure and how you can leverage it for your workloads.
Google is a leader in artificial intelligence with the invention of technologies like TensorFlow. Did you know you can leverage Google’s technology for your own projects? Learn about Google's history of innovation in AI infrastructure and how you can leverage it for your workloads.
Common Uses
Run large-scale AI training
Powerful, scalable, and efficient AI training
The AI Hypercomputer architecture offers optionality to use the underlying infrastructure that best scales to meet your training needs.
Powerful, scalable, and efficient AI training
Measure the effectiveness of your large scale training the Google way with ML Productivity Goodput.
Character AI leverages Google Cloud to scale up
"We need GPUs to generate responses to users' messages. And as we get more users on our platform, we need more GPUs to serve them. So on Google Cloud, we can experiment to find what is the right platform for a particular workload. It's great to have that flexibility to choose which solutions are most valuable." Myle Ott, Founding Engineer, Character.AI
How-tos
Powerful, scalable, and efficient AI training
The AI Hypercomputer architecture offers optionality to use the underlying infrastructure that best scales to meet your training needs.
Additional resources
Powerful, scalable, and efficient AI training
Measure the effectiveness of your large scale training the Google way with ML Productivity Goodput.
Customer examples
Character AI leverages Google Cloud to scale up
"We need GPUs to generate responses to users' messages. And as we get more users on our platform, we need more GPUs to serve them. So on Google Cloud, we can experiment to find what is the right platform for a particular workload. It's great to have that flexibility to choose which solutions are most valuable." Myle Ott, Founding Engineer, Character.AI
Deliver AI powered applications
Leverage open frameworks to deliver AI powered experiences
Google cloud is committed to ensuring open frameworks work well within the AI Hypercomputer architecture.
Leverage open frameworks to deliver AI powered experiences
Google Cloud's open software ecosystem allows you to build applications with the tools and frameworks you are most comfortable with, while taking advantage of the price-performance benefits of the AI Hypercomputer architecture.
Priceline: Helping travelers curate unique experiences
"Working with Google Cloud to incorporate generative AI allows us to create a bespoke travel concierge within our chatbot. We want our customers to go beyond planning a trip and help them curate their unique travel experience." Martin Brodbeck, CTO, Priceline
How-tos
Leverage open frameworks to deliver AI powered experiences
Google cloud is committed to ensuring open frameworks work well within the AI Hypercomputer architecture.
Additional resources
Leverage open frameworks to deliver AI powered experiences
Google Cloud's open software ecosystem allows you to build applications with the tools and frameworks you are most comfortable with, while taking advantage of the price-performance benefits of the AI Hypercomputer architecture.
Customer examples
Priceline: Helping travelers curate unique experiences
"Working with Google Cloud to incorporate generative AI allows us to create a bespoke travel concierge within our chatbot. We want our customers to go beyond planning a trip and help them curate their unique travel experience." Martin Brodbeck, CTO, Priceline
Cost efficiently serve models at scale
Maximize price/performance for serving AI at scale
Google Cloud provides industry leading price/performance for serving AI models with accelerator optionality to address any workload's needs.
Maximize price/performance for serving AI at scale
Cloud TPU v5e and G2 VM Instances delivering NVIDIA L4 GPUs enable high-performance and cost-effective inference for a wide range of AI workloads, including the latest LLMs and Gen AI models. Both offer significant price performance improvements over previous models and Google Cloud's AI Hypercomputer architecture enables customers to scale their deployments to industry leading levels.
Assembly AI leverage Google Cloud for cost efficiency
"Our experimental results show that Cloud TPU v5e is the most cost-efficient accelerator on which to run large-scale inference for our model. It delivers 2.7x greater performance per dollar than G2 and 4.2x greater performance per dollar than A2 instances." Domenic Donato,
VP of Technology, AssemblyAI
How-tos
Maximize price/performance for serving AI at scale
Google Cloud provides industry leading price/performance for serving AI models with accelerator optionality to address any workload's needs.
Additional resources
Maximize price/performance for serving AI at scale
Cloud TPU v5e and G2 VM Instances delivering NVIDIA L4 GPUs enable high-performance and cost-effective inference for a wide range of AI workloads, including the latest LLMs and Gen AI models. Both offer significant price performance improvements over previous models and Google Cloud's AI Hypercomputer architecture enables customers to scale their deployments to industry leading levels.
Customer examples
Assembly AI leverage Google Cloud for cost efficiency
"Our experimental results show that Cloud TPU v5e is the most cost-efficient accelerator on which to run large-scale inference for our model. It delivers 2.7x greater performance per dollar than G2 and 4.2x greater performance per dollar than A2 instances." Domenic Donato,
VP of Technology, AssemblyAI
Open source models on Google Cloud
Serve a model with GKE on a single GPU
Train common models with GPUs
Scale model serving to multiple GPUs
Serve an LLM using multi-host TPUs on GKE with Saxml
Train at scale with the NVIDIA Nemo framework
- English
- Deutsch
- Español
- Español (Latinoamérica)
- Français
- Indonesia
- Italiano
- Português (Brasil)
- 简体中文
- 繁體中文
- 日本語
- 한국어
- Accelerate your digital transformation
- Whether your business is early in its journey or well on its way to digital transformation, Google Cloud can help solve your toughest challenges.
- Key benefits
- Featured Products
- AI and Machine Learning
- Business Intelligence
- Compute
- Containers
- Data Analytics
- Databases
- Developer Tools
- Distributed Cloud
- Hybrid and Multicloud
- Industry Specific
- Integration Services
- Management Tools
- Maps and Geospatial
- Media Services
- Migration
- Mixed Reality
- Networking
- Operations
- Productivity and Collaboration
- Security and Identity
- Serverless
- Storage
- Web3
- Featured Products
- Not seeing what you're looking for?
- See all products (100+)
- Not seeing what you're looking for?
- See all AI and machine learning products
- Business Intelligence
- Not seeing what you're looking for?
- See all compute products
- Not seeing what you're looking for?
- See all data analytics products
- Not seeing what you're looking for?
- See all developer tools
- Hybrid and Multicloud
- Industry Specific
- Not seeing what you're looking for?
- See all management tools
- Media Services
- Not seeing what you're looking for?
- See all networking products
- Productivity and Collaboration
- Not seeing what you're looking for?
- See all security and identity products
- Save money with our transparent approach to pricing
- Google Cloud's pay-as-you-go pricing offers automatic savings based on monthly usage and discounted rates for prepaid resources. Contact us today to get a quote.
- Pricing overview and tools
- Product-specific Pricing
- Learn & build
- Connect
- Accelerate your digital transformation
- Learn more
- Key benefits
- Why Google Cloud
- AI and ML
- Multicloud
- Global infrastructure
- Data Cloud
- Modern Infrastructure Cloud
- Security
- Productivity and collaboration
- Reports and insights
- Executive insights
- Analyst reports
- Whitepapers
- Customer stories
- Industry Solutions
- Retail
- Consumer Packaged Goods
- Financial Services
- Healthcare and Life Sciences
- Media and Entertainment
- Telecommunications
- Games
- Manufacturing
- Supply Chain and Logistics
- Government
- Education
- See all industry solutions
- See all solutions
- Application Modernization
- CAMP
- Modernize Traditional Applications
- Migrate from PaaS: Cloud Foundry, Openshift
- Migrate from Mainframe
- Modernize Software Delivery
- DevOps Best Practices
- SRE Principles
- Day 2 Operations for GKE
- FinOps and Optimization of GKE
- Run Applications at the Edge
- Architect for Multicloud
- Go Serverless
- Artificial Intelligence
- Customer Engagement Suite with Google AI
- Document AI
- Vertex AI Search for retail
- Gemini for Google Cloud
- Generative AI on Google Cloud
- APIs and Applications
- New Business Channels Using APIs
- Unlocking Legacy Applications Using APIs
- Open Banking APIx
- Data Analytics
- Data Migration
- Data Lake Modernization
- Stream Analytics
- Marketing Analytics
- Datasets
- Business Intelligence
- AI for Data Analytics
- Databases
- Database Migration
- Database Modernization
- Databases for Games
- Google Cloud Databases
- Migrate Oracle workloads to Google Cloud
- Open Source Databases
- SQL Server on Google Cloud
- Gemini for Databases
- Infrastructure Modernization
- Application Migration
- SAP on Google Cloud
- High Performance Computing
- Windows on Google Cloud
- Data Center Migration
- Active Assist
- Virtual Desktops
- Rapid Migration and Modernization Program
- Backup and Disaster Recovery
- Red Hat on Google Cloud
- Cross-Cloud Network
- Observability
- Productivity and Collaboration
- Google Workspace
- Google Workspace Essentials
- Cloud Identity
- Chrome Enterprise
- Cloud Search
- Security
- Security Analytics and Operations
- Web App and API Protection
- Security and Resilience Framework
- Risk and compliance as code (RCaC)
- Software Supply Chain Security
- Security Foundation
- Google Cloud Cybershield™
- Startups and SMB
- Startup Program
- Small and Medium Business
- Software as a Service
- Featured Products
- Compute Engine
- Cloud Storage
- BigQuery
- Cloud Run
- Google Kubernetes Engine
- Vertex AI
- Looker
- Apigee API Management
- Cloud SQL
- Gemini
- Cloud CDN
- See all products (100+)
- AI and Machine Learning
- Vertex AI Platform
- Vertex AI Studio
- Vertex AI Agent Builder
- Conversational Agents
- Vertex AI Search
- Speech-to-Text
- Text-to-Speech
- Translation AI
- Document AI
- Vision AI
- Contact Center as a Service
- See all AI and machine learning products
- Business Intelligence
- Looker
- Looker Studio
- Compute
- Compute Engine
- App Engine
- Cloud GPUs
- Migrate to Virtual Machines
- Spot VMs
- Batch
- Sole-Tenant Nodes
- Bare Metal
- Recommender
- VMware Engine
- Cloud Run
- See all compute products
- Containers
- Google Kubernetes Engine
- Cloud Run
- Cloud Build
- Artifact Registry
- Cloud Code
- Cloud Deploy
- Migrate to Containers
- Deep Learning Containers
- Knative
- Data Analytics
- BigQuery
- Looker
- Dataflow
- Pub/Sub
- Dataproc
- Cloud Data Fusion
- Cloud Composer
- BigLake
- Dataplex
- Dataform
- Analytics Hub
- See all data analytics products
- Databases
- AlloyDB for PostgreSQL
- Cloud SQL
- Firestore
- Spanner
- Bigtable
- Datastream
- Database Migration Service
- Bare Metal Solution
- Memorystore
- Developer Tools
- Artifact Registry
- Cloud Code
- Cloud Build
- Cloud Deploy
- Cloud Deployment Manager
- Cloud SDK
- Cloud Scheduler
- Cloud Source Repositories
- Infrastructure Manager
- Cloud Workstations
- Gemini Code Assist
- See all developer tools
- Distributed Cloud
- Google Distributed Cloud Connected
- Google Distributed Cloud Air-gapped
- Hybrid and Multicloud
- Google Kubernetes Engine
- Apigee API Management
- Migrate to Containers
- Traffic Director
- Cloud Build
- Operations
- Google Distributed Cloud
- Industry Specific
- Anti Money Laundering AI
- Cloud Healthcare API
- Device Connect for Fitbit
- Telecom Network Automation
- Telecom Data Fabric
- Telecom Subscriber Insights
- Spectrum Access System (SAS)
- Integration Services
- Application Integration
- Workflows
- Apigee API Management
- Cloud Tasks
- Cloud Scheduler
- Dataproc
- Cloud Data Fusion
- Cloud Composer
- Pub/Sub
- Eventarc
- Management Tools
- Cloud Shell
- Cloud console
- Cloud Endpoints
- Cloud IAM
- Cloud APIs
- Service Catalog
- Cost Management
- Operations
- Carbon Footprint
- Config Connector
- Active Assist
- See all management tools
- Maps and Geospatial
- Earth Engine
- Google Maps Platform
- Media Services
- Cloud CDN
- Live Stream API
- OpenCue
- Transcoder API
- Video Stitcher API
- Migration
- Migration Center
- Application Migration
- Migrate to Virtual Machines
- Cloud Foundation Toolkit
- Database Migration Service
- Migrate to Containers
- BigQuery Data Transfer Service
- Rapid Migration and Modernization Program
- Transfer Appliance
- Storage Transfer Service
- VMware Engine
- Mixed Reality
- Immersive Stream for XR
- Networking
- Cloud Armor
- Cloud CDN and Media CDN
- Cloud DNS
- Cloud Load Balancing
- Cloud NAT
- Cloud Connectivity
- Network Connectivity Center
- Network Intelligence Center
- Network Service Tiers
- Virtual Private Cloud
- Private Service Connect
- See all networking products
- Operations
- Cloud Logging
- Cloud Monitoring
- Error Reporting
- Cloud Debugger
- Cloud Trace
- Cloud Profiler
- Cloud Quotas
- Productivity and Collaboration
- AppSheet
- AppSheet Automation
- Google Workspace
- Google Workspace Essentials
- Gemini for Workspace
- Cloud Identity
- Chrome Enterprise
- Security and Identity
- Cloud IAM
- Sensitive Data Protection
- Mandiant Managed Defense
- Google Threat Intelligence
- Security Command Center
- Cloud Key Management
- Mandiant Incident Response
- Chrome Enterprise Premium
- Assured Workloads
- Google Security Operations
- Mandiant Consulting
- See all security and identity products
- Serverless
- Cloud Run
- Cloud Functions
- App Engine
- Workflows
- API Gateway
- Storage
- Cloud Storage
- Block Storage
- Filestore
- Persistent Disk
- Cloud Storage for Firebase
- Local SSD
- Storage Transfer Service
- Parallelstore
- Google Cloud NetApp Volumes
- Backup and DR Service
- Web3
- Blockchain Node Engine
- Blockchain RPC
- Save money with our transparent approach to pricing
- Request a quote
- Pricing overview and tools
- Google Cloud pricing
- Pricing calculator
- Google Cloud free tier
- Cost optimization framework
- Cost management tools
- Product-specific Pricing
- Compute Engine
- Cloud SQL
- Google Kubernetes Engine
- Cloud Storage
- BigQuery
- See full price list with 100+ products
- Learn & build
- Google Cloud Free Program
- Quickstarts
- Cloud computing basics
- Blog
- Learning Hub
- Training
- Certification
- Cloud Architecture Center
- Connect
- Innovators
- Developer Center
- Events and webinars
- Google Cloud Community
- Google Cloud Consulting
- Google Cloud Consulting
- Google Cloud Marketplace
- Google Cloud partners
- Become a partner
