Skip to content

Commit

Permalink
fix figures in README
Browse files Browse the repository at this point in the history
  • Loading branch information
zhizhenzhong committed Oct 6, 2023
1 parent d7f97c1 commit 3754de8
Show file tree
Hide file tree
Showing 2 changed files with 6 additions and 5 deletions.
2 changes: 1 addition & 1 deletion kit/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -5,7 +5,7 @@
The Lightning developer kit contains the minimum-viable photonic devices to perform a single multiplication operation in the photonic domain. The current version of Lightning developer kit takes in two input signals (in analog voltages, represented by $A$ and $B$), and returns an output signal (in analog voltages, represented by $R$). For each operation, $R=A*B$ is performed in photonic domain inside the Lightning developer kit. For more details, please refer to the Appendix of our [paper](https://doi.org/10.1145/3603269.3604821).

<p align="center">
<img src="graphics/devkitV2.png" alt="kit" width=70% height=auto>
<img src="../graphics/devkitV2.png" alt="kit" width=90% height=auto>
</p>

## Shopping list
Expand Down
9 changes: 5 additions & 4 deletions rtl/README.md
Original file line number Diff line number Diff line change
@@ -1,6 +1,7 @@
# Lightning Register Transfer Level (RTL) Code
The Lightning RTL code implements the fast datapath
The code in this folder reproduces results reported in Lightning paper Section 6.
The Lightning RTL code implements the novel "reconfigurable count-action" based datapath that handles the data movement from ML inference request packets to the final result packets. In particular, our reconfigurable count-action abstraction decouples the control and data planes of inference requests by enabling the datapath to keep track of the directed acyclic graph (DAG) of each inference request for DNN models without interrupting the flow of the data in and out of photonic computing cores.

The code in this folder reproduces results reported in Lightning paper Section 6.

## Reproduce the testbench results

Expand All @@ -9,11 +10,11 @@ Run ```python tb/create_venv.py``` to create the environments and dependencies<b
Run ```sudo apt install -y verilator python3 python3-pip python3-venv``` to install verilator<br>
Run ```verilator --version``` to check if the install version is ```Verilator 4.038 2020-07-11 rev v4.036-114-g0cd4a57ad```. Other Verilator versions may or may not be compatible due to frequent Verilator codebase changes.

### 2. Build the verilator testbench
### 2. Build the Verilator testbench
Run ```make build-sw-lenet-single-core``` <br>
If build is successful, the compiled Verilator testbench will appear in the ```obj_dir/``` folder.

### 3. Run the verilator testbench (e.g., LeNet-300-100 DNN)
### 3. Run the Verilator testbench (e.g., LeNet-300-100 DNN)
Run ```make run-sw-lenet-single-core``` <br>
After running the testbench, results will appear in the ```tb/``` folder:
- ```tb_lenet_sim.vcd```: the generated waveform file containing all registers values over the simulation period.
Expand Down

0 comments on commit 3754de8

Please sign in to comment.