What is TPM? · Ultra Transcenders glossary

Tokens per minute, the measure for Azure OpenAI quota and deployment rate limits. Going over it produces an HTTP 429 response.

Also called tokens per minute.

In the Ultra Transcenders books

Each book explains TPM in context, with comparison tables and the common traps.

Measure
A figure like total revenue, worked out by aggregating fact table data wherever dimensions intersect within a semantic model.
Azure OpenAI
Use of OpenAI models in Azure, including GPT, o-series, embeddings, image and Whisper, through either a Foundry resource or an Azure OpenAI resource. Content Understanding isn't available from a standalone Azure OpenAI resource.
Model deployment
Inside a resource, each deployment is a named copy of a Foundry model with its own TPM quota and deployment type. Requests identify the model by this deployment name.
OVER
Gives a T-SQL window function its window: PARTITION BY, ORDER BY and, if wanted, a ROWS or RANGE frame. Rankings and running totals can then be worked out while every row is kept.
HTTP 429
The HTTP status a service sends when it is throttling a caller, for example RequestRateTooLarge in Cosmos DB. Callers should wait for the period given in the retry-after header before retrying.

Confidential disk encryption
Available just for confidential VMs, this encrypts the OS disk with keys bound to that VM's TPM, meaning the disk is unreadable anywhere else.
HTTP 429 (Azure OpenAI)
Status returned once a deployment goes over its RPM or TPM limit, or when capacity is throttled. Back off exponentially and respect retry-after-ms before retrying, because failed calls still count against the limit.
Network Unlock
With this BitLocker capability, domain-joined machines secured by TPM plus PIN can boot without the PIN when connected by cable inside the company. A certificate and a WDS server running Network Unlock are prerequisites.
PIN
A brief secret code. The Windows Hello for Business kind is bound to a single machine and protected by its TPM, which means it stays put rather than travelling like a password does. App protection policies may additionally ask for a separate app-level code.
Self-deploying mode
Joins a device to Microsoft Entra with no user input at all, relying on TPM attestation; this Windows Autopilot mode suits kiosks and shared devices.
Silent encryption
Turning on BitLocker through Intune with no prompts for the user, with the recovery key stored in Microsoft Entra ID. Devices need a TPM and compatible hardware for it to work.