TT-Topology

Tenstorrent Topology (TT-Topology) is a command line utility used to flash multiple n150 or n300 cards on a system to use specific single-host ETH routing configurations.

It currently supports three configurations: mesh, linear, and torus.

Warning

tt-topology is not designed to be used with the following products:

• BH PCIe cards
• WH 6U Galaxy systems
• BH 6U Galaxy systems

The tool will throw an error if used with unsupported products.

Additionally, tt-topology is designed to be used only in a single-host context. Multi-host topologies will not be discovered.

Official Repository

https://github.com/tenstorrent/tt-topology/

Getting started

Install Rust (if you don't already have it)

If Rust isn't already installed on your system, you can install it through either of the following methods:

Using Distribution packages (preferred)

• Fedora / EL9

  sudo dnf install cargo

• Ubuntu / Debian

  sudo apt install cargo

Using Rustup

curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
source "$HOME/.cargo/env"

Installation (for users)

tt-topology is available on PyPi and can be installed in your Python (v3.10 and up) environment using pip.

pip install tt-topology

Important

It is always recommended to manage, build, and install Python packages within a virtual environment.

A virtual environment can be created using venv:

python -m venv .venv
source .venv/bin/activate

Installation (for developers)

Clone the repository

git clone https://github.com/tenstorrent/tt-topology.git
cd tt-topology

Install

pip install .

or for users who would like to edit the code without re-building, install tt-topology in editable mode.

pip install --editable .

Recommended: install the pre-commit hooks so there is auto formatting for all files on committing.

pre-commit install

Usage

Command line arguments

usage: tt-topology [-h] [-v] [-l {linear,torus,mesh,isolated}] [-o] [-f [filename]] [-g] [-ls] [--log [log]] [-p [plot]] [-r [config.json ...]]

Tenstorrent Topology (TT-Topology) is a command line utility to flash ethernet coordinates when multiple NB's are connected together.

options:
  -h, --help            show this help message and exit
  -v, --version         show program's version number and exit
  -l {linear,torus,mesh,isolated}, --layout {linear,torus,mesh,isolated}
                        Select the layout (linear, torus, mesh, isolated). Default is linear.
  -o, --octopus
  -f [filename], --filename [filename]
                        Change filename for test log. Default: ~/tt_smi/<timestamp>_snapshot.json
  -g, --generate_reset_json
                        Generate default reset json file that reset consumes. Update the generated file and use it as an input for the --reset option
  -ls, --list           List out all the boards on host with their coordinates and layout.
  --log [log]           Change filename for the topology flash log. Default: ~/tt_topology_logs/<timestamp>_log.json
  -p [plot], --plot_filename [plot]
                        Change the plot of the png that will have the graph layout of the chips. Default: chip_layout.png
  -r [config.json ...], --reset [config.json ...]
                        Provide a valid reset JSON

TT-Topology Procedure

TT-Topology does the following when calculating and flashing the coordinates:

1. Flash all the boards to default - set all eth port disables to 0 and reset coordinates to (0,0) for local chips and (1,0) for n300 remote chips.
2. Issue a board level reset to apply the new flash to the chips.
3. Generate a mapping of all possible connections and their type between the available chips.
4. Using a graph algorithm generate coordinates for each chip based on user input. These layouts are discussed in detail in the sections below.
5. Write the new coordinates to the chips.
6. Issue a board level reset to apply the new flash to the chips.
7. Return a png with a graphic representation of the layout and a .json log file with details of the above steps.

Chip layouts

TT-Topology can be used to flash one of the three chip layouts: mesh, linear and torus.

Mesh

The mesh layout is a trivalent graph where each node can have a maximum of three connections. A BFS algorithm is used to assign the coordinates.

The command to generate a mesh layout is:

$ tt-topology -l mesh -p mesh_layout.png

For a host with two n300 cards and four n300 cards, the command will generate layouts that look like the following:

   

Linear

The linear layout, as the name suggests, is a layout where all chips are connected in a single line. The coordinates are assigned by finding a cycle in the graph and then assigning coordinates in order.

The command to generate a linear layout is:

$ tt-topology -l linear -p linear_layout.png

For a host with two n300 cards and four n300 cards, the command will generate layouts that look like the following:

   

Torus

The torus layout is a cyclic graph where a single line connects all nodes.

The coordinates are assigned by finding a cycle in the graph and then assigning coordinates in order.

The command to generate a torus layout is:

$ tt-topology -l torus -p torus_layout.png

For a host with two n300 cards and four n300 cards, the command will generate layouts that look like the following:

   

Octopus (TGG/TG) Support in TT-Topology

• TGG setting: 8 n150 cards connected to 2 Galaxy 4U systems
• TG setting: 8 n150 cards connected to 1 Galaxy 4U system

Usage

1. Generate a default mobo reset json file saved at ~/.config/tenstorrent/reset_config.json by running the following command

   $ tt-topology -g
   

2. Fill in "mobo", "credo", and "disabled_ports" under "wh_mobo_reset"

   Here is an example of what your reset_config.json file may look like:

   {
       "time": "2024-03-06T20:12:27.640859",
       "host_name": "yyz-lab-212",
       "gs_tensix_reset": {
           "pci_index": []
       },
       "wh_link_reset": {
           "pci_index": [
               0,
               1,
               2,
               3
           ]
       },
       "re_init_devices": true,
       "wh_mobo_reset": [
           {
               "nb_host_pci_idx": [
                   0,
                   1,
                   2,
                   3
                 ],
               "mobo": "mobo-ce-44",
               "credo": [
                   "6:0",
                   "6:1",
                   "7:0",
                   "7:1"
               ],
               "disabled_ports": [
                   "0:2",
                   "1:2",
                   "6:2",
                   "7:2"
               ]
           }
       ]
   }
   

3. Flashing multiple NB cards to use specific eth routing configurations by running the following command

   $ tt-topology -o -r ~/.config/tenstorrent/reset_config.json
   

Internal Procedure

1. Setup mobo_eth_en on every local n150 to train with the Galaxy
2. Program the shelf/rack of the Galaxies
3. Program all local n150s to rack 0, shelf 0, x 0, y 0
4. Reset with the following retimer_sel and disable_sel and wait for training
   • retimer_sel: From the credo field of the reset json file for the specific Galaxy
   • disable_sel: All the other ports not specified by the retimer_sel
5. Check QSFP link and change shelf number for each n150 according to the shelf on the connected Galaxy
6. Program the x, y coords of the local n150s based on the other side of the link
7. Reset again with the retimer_sel and disable_sel and wait for training, and verify all chips show up
   • retimer_sel: From the credo field of the reset json file for the specific Galaxy
   • disable_sel: From the disabled_ports field of the reset json file for the specific Galaxy

Logging

TT-Topology records the pre- and post-flash relevant SPI parameter values, connection maps, and coordinates of the chips in a .json file for record keeping and debugging. By default it is stored at ~/tt_topology_logs/<timestamp>_log.json. This can be changed by using the --log CLI arg:

$ tt-topology --log new_log.json ...

License

Apache 2.0 - https://www.apache.org/licenses/LICENSE-2.0.txt