(tpu7x-intro)=
# {{ product_tpu7x }}

{{ product_tpu7x }} is the generation for the largest training and inference workloads. Each chip has {{ tpu7x_tensorcores }} TensorCores and {{ tpu7x_sparsecores }} SparseCores, and a pod holds up to {{ tpu7x_chips_per_pod }} chips.

Use the specification topic to size a slice, the configurations topic to choose a slice shape, and the VM types topic to see what each VM provides.

```{include} ../_includes/tpu-warehouse--note-tpu-api-tpu7x.md
```

```{include} ../_includes/tpu-warehouse--note-tensorflow-tpu7x.md
```


:::::{container} manifest-nav

-   **[{{ product_tpu7x }} specifications](../topics/tpu7x-specifications.md)**  
Per-chip and per-pod specifications for {{ product_tpu7x }}.
-   **[{{ product_tpu7x }} supported configurations](../topics/tpu7x-configurations.md)**  
The slice shapes you can request for {{ product_tpu7x }}, and how many chips and hosts each uses.
-   **[{{ product_tpu7x }} VM types](../topics/tpu7x-vm-types.md)**  
The VM types available for {{ product_tpu7x }} and the host resources each one provides.
% manifest:published-in unfiltered,tpu7x
-   **[{{ product_tpu7x }} memory hierarchy](../topics/tpu7x-memory.md)**  
How high-bandwidth memory \(HBM\) and host memory are organized on a {{ product_tpu7x }} chip.
% manifest:end

:::::



:::::{container} manifest-nav

**Related information**  


[{{ product_tpu7x }} specifications](../topics/tpu7x-specifications.md)

[{{ product_tpu7x }} supported configurations](../topics/tpu7x-configurations.md)

[{{ product_tpu7x }} VM types](../topics/tpu7x-vm-types.md)

[Create a TPU VM](../topics/tpu-create-vm.md)

:::::

```{toctree}
:maxdepth: 2
:hidden:

tpu7x-specifications
tpu7x-configurations
tpu7x-vm-types
tpu7x-memory
```