Teradata Cloud is Teradata’s managed data analytics service on AWS, Azure and Google Cloud. Teradata provisions, monitors and manages the environment in a Teradata-owned account or in the customer’s own cloud account, and customises the configuration to the subscription.[1] It was formerly called VantageCloud Enterprise.[2] The documentation describes two compute pricing models. Blended pricing sells license tiers (Base, Advanced, Enterprise) and meters compute in TCore-Hours. Unit consumption pricing sells packages (Teradata Cloud and Teradata Cloud +) and meters compute in Units derived from logical I/O.[1][7] Teradata’s current pricing page describes Fixed + Flex Pricing in Teradata Units, which combines a committed baseline with elastic capacity in the same way.[14] See Teradata licensing for the overview.
Editions
License tiers (blended pricing). The tier table lists Premier Cloud support for every tier.[3] The Teradata Cloud version of the table labels its third column “Teradata Cloud” instead of “Enterprise”, apparently because of the rename.[4]
| Feature | Base | Advanced | Enterprise |
|---|---|---|---|
| Elastic Performance on Demand (blended pricing option required) | No | Yes | Yes |
| Unit consumption (applicable as-a-service deployments only) | No | No | Yes |
| Teradata Intelligent Memory | No | Yes | Yes |
| Workload management | None listed | TIWM | TASM |
| Columnar, Native Object Store, Row-Level Security, Secure Zones, Temporal, block level compression | Yes | Yes | Yes |
Source: VantageCloud Enterprise license tier table.[3]
Packages (unit consumption pricing). Both packages list “Compute Metric: Units”.[5] The Teradata Cloud Package includes Premier Cloud support and a USD 10,000 Teradata Education Checkbook. The Teradata Cloud + Package adds an Industry Data Model or Industry Analytic Schema, Premier Cloud + Priority Service, a USD 25,000 Education Checkbook and further advisory services.[5] Both packages include Viewpoint, PDCR, Data Lab, Data Mover, Teradata Fabric and Query Service.[5] In the VantageCloud Enterprise documentation, the same packages are called Enterprise and Enterprise +.
Metrics
| Teradata term | Catalog row | Definition |
|---|---|---|
| TCore-Hour | TCore-Hour | “A TCore-Hour is a metric representing the availability of compute and I/O resources on the underlying infrastructure, by the hour.”[9] |
| Unit | Unit (Teradata Cloud) | Consumed “when Teradata Cloud compute resources are active and available to perform workloads”.[6] |
| Storage | Terabyte of storage | Block storage is ordered per system and can only be increased.[7] |
Counting / floors
TCore ratings. Each instance type has a TCore rating. Under blended pricing, the draw of TCore-Hours depends on “the number, type, and TCore ratings of provisioned instances”.[9] On AWS, for example, an R7i XSmall instance is rated 1.52 TCore, a Medium 6.06, an XLarge 12.11 and a 3XLarge 30.63. Older R5 and R5b series have their own ratings.[8] A system can have between 2 and 128 instances.[8] As commentary: at a given rating, a system’s hourly TCore-Hour draw is roughly the sum of its instances’ ratings while it runs. Teradata publishes no formula beyond the factors listed. Catalog: TCore-Hour draw depends on instance number, type and TCore rating.
Unit formula. Under unit consumption pricing, hourly consumption is calculated as “(Total Logical IO TB Used / (Max LIO TB Rate per Instance * No. of Instance)) * Max Unit rate”.[6] Teradata publishes, for each instance size, a maximum logical I/O rate and a minimum and maximum unit rate per instance per hour:
| Size | Max LIO (TB/hour) | Teradata Cloud Units (min / max) | Teradata Cloud + Units (min / max) |
|---|---|---|---|
| S | 8 | 2.5 / 10 | 3 / 12 |
| M | 12 | 5 / 20 | 6 / 24 |
| L | 16 | 7.5 / 30 | 9 / 36 |
| XL | 24 | 10 / 40 | 12 / 48 |
| 2XL | 32 | 15 / 60 | 18 / 72 |
| 3XL | 64 | 20 / 80 | 24 / 96 |
Source: Teradata Cloud Admin Console documentation. Availability of 2XL and 3XL varies by cloud provider.[6] The minimum rate acts as a floor. An idle but running S instance on the Teradata Cloud package still consumes 2.5 Units per hour. Logical I/O counts exclude collector queries, Teradata system-user activity, PDCR data movement and PDCR admin SQL, and user queries stopped by a processing error.[6] Catalog: Teradata Cloud Units are calculated from logical I/O; Teradata Cloud logical I/O excludes system and failed work.
Commitments. The Admin Console compares actual usage with committed usage for the term. Consumption data takes 72 hours to appear.[6] Catalog: Teradata Cloud consumption data appears after 72 hours.
Virtualization & partitioning
Teradata Cloud has no core-based partitioning. Capacity changes through documented elasticity operations, each with its own pricing effect.[7]
| Operation | Effect on consumption | Availability |
|---|---|---|
| Stop/Start | Pauses or resumes TCore-Hour consumption | Flexible Capacity compute pricing option only[7] |
| Scale Out/In | Changes the TCore-Hour or Unit rate with the number and size of instances | Blended pricing and unit consumption; change request needed for unit consumption[7] |
| Scale Up/Down | Changes the TCore-Hour or Unit rate with node instance size | As above[7] |
| Elastic Performance on Demand (EPOD) | Adds compute without downtime, up to the baseline system size | Blended pricing only[7] |
| Block storage expansion | New order; one-way | AWS 1 TB, Azure 1 TiB per node, Google Cloud 1 TB increments[7] |
Teradata Cloud documentation refers to the Cloud Service Description, knowledge article KB0047025, for which elasticity features apply to each pricing model and cloud provider.[7] In May 2026 Teradata added elastic compute, on-demand capacity for training, experimentation and bursts, which runs next to always-on active compute in one system.[13] Catalog: Stop/Start pauses TCore-Hour consumption (Flexible Capacity only); EPOD adds compute up to the baseline size, blended pricing only; Unit-priced systems scale through a Change Request; Teradata Cloud block storage expansion is one-way.
Cloud / BYOL
Customers subscribe directly with Teradata or through a private offer, and Teradata works with them on the details of the offer.[16] Teradata Cloud 3.4 can be deployed on AWS, Azure and Google Cloud. The do-it-yourself (DIY) deployment option supports only Azure and AWS.[16] An earlier route sold Teradata Vantage on AWS through AWS Marketplace as an annual quantity of TCore-Hours. Customers estimated the quantity with Teradata’s TCore-Hours calculator and could buy more “at any time, at the same price, before your subscription expires”.[10] The cloud infrastructure cost is part of the managed service. No published right lets on-premises Teradata licences be converted into Teradata Cloud capacity. Catalog: AWS Marketplace TCore-Hours can be topped up at the same price before expiry.
Programs
- Blended Pricing: Fixed Capacity and Flexible Capacity compute pricing options metered in TCore-Hours.[7]
- Unit Consumption Pricing Model: packages metered in Units against committed usage.[5][6]
- Teradata Vantage on AWS Marketplace: annual TCore-Hour subscriptions, documented in a 2021 guide.[10]
- Vantage Trial: 10,000 free TCore-Hours over 30 days, documented in a 2022 guide.[9]
Support and service level. Teradata Cloud includes operations by Teradata: patches, upgrades, backups, performance reports and support tickets. Teradata supports 99.9% service availability for Teradata Database, excluding scheduled maintenance.[12] The Product Support Policies do not cover VantageCloud services. Those are covered in the applicable Teradata Cloud Description document.[15]
Audits and compliance
Telemetry, not an audit, is the main compliance mechanism. Under the Teradata Master Agreement, telemetry collection agents send data to Teradata to detect and resolve issues, analyse and report system usage, and “Enable metering and billing capabilities”. Teradata states that it does not collect personally identifiable information or customer data stored on its systems.[11] Customers can download consumption data from the Admin Console and view it hourly, daily or monthly.[6] A licence manager should reconcile the downloaded data with invoices and with the committed quantity in the order. Catalog: Telemetry collected under the Master Agreement drives metering and billing.
Out of scope
- Prices for Teradata Cloud tiers and packages, which Teradata does not publish.
- The Teradata Cloud Service Description (KB0047025), order forms and the Master Agreement.
- Azure and Google Cloud TCore ratings in detail. They are published on separate pages with the same structure as the AWS page.
- VantageCloud Lake, covered in Teradata VantageCloud Lake licensing, and on-premises licensing, covered in Teradata on-premises, VantageCore and support.