Registrati per l'accesso sperimentale all'SDK Google Tensor ML
L'SDK Google Tensor ML ti consente di creare funzionalità di machine learning on-device specificamente per gli smartphone Google Pixel, sfruttando al contempo il SoC (System on a Chip) Tensor progettato su misura di Google Pixel con il suo acceleratore di inferenza Tensor Processing Unit (TPU) dedicato.
L'accesso sperimentale consente agli sviluppatori di provare a utilizzare l'SDK Google Tensor ML prima della disponibilità generale.
[[["Facile da capire","easyToUnderstand","thumb-up"],["Il problema è stato risolto","solvedMyProblem","thumb-up"],["Altra","otherUp","thumb-up"]],[["Mancano le informazioni di cui ho bisogno","missingTheInformationINeed","thumb-down"],["Troppo complicato/troppi passaggi","tooComplicatedTooManySteps","thumb-down"],["Obsoleti","outOfDate","thumb-down"],["Problema di traduzione","translationIssue","thumb-down"],["Problema relativo a esempi/codice","samplesCodeIssue","thumb-down"],["Altra","otherDown","thumb-down"]],["Ultimo aggiornamento 2025-07-24 UTC."],[],[],null,["# Sign up for experimental access to Google Tensor ML SDK\n\nGoogle Tensor ML SDK empowers you to build on-device machine learning\ncapabilities specifically for Google Pixel phones, while leveraging Google\nPixel's custom-designed Tensor System-on-Chip (SoC) with its dedicated Tensor\nProcessing Unit (TPU) inference accelerator.\n\nExperimental access is for developers to try using Google Tensor ML SDK before\ngeneral availability.\n| **Note:** By signing up for experimental access, you agree to these [terms](/edge/litert/next/tensor_ml_terms) that apply to your use of Google Tensor ML SDK.\n\n[Sign up for experimental\naccess](https://services.google.com/fb/forms/tensor_ml_sdk_experimental_access)\n\nKey features\n------------\n\n- **Access a curated collection of open-source, SDK-compatible models** using Model Garden.\n- **Visualize model architecture** for better understanding of on-device performance.\n- **Ensure TPU compatibility** by checking operation support within your models.\n- **Analyze the estimated latency** of individual operations on the TPU.\n- **Profile inference performance** by comparing CPU and TPU execution on a Pixel device.\n\n| **Note:** There is no guarantee that a feature in experimental access will ever become generally available. It may remain an experimental access feature indefinitely, be shutdown, or be combined with other features into packages designed for specific developer audiences. Some experimental access features may eventually become generally available, but there is no fixed release schedule."]]