-
Notifications
You must be signed in to change notification settings - Fork 12
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
1 changed file
with
19 additions
and
15 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,4 +1,9 @@ | ||
| # Stella Nera - Halutmatmul | ||
| <div align="center"> | ||
|
|
||
| # Stella Nera: A halutmatmul based accelerator | ||
| </div> | ||
|
|
||
| <div align="center"> | ||
|
|
||
| ### Algorithmic CI | ||
| [](https://github.com/joennlae/halutmatmul/actions/workflows/python_testing.yaml) | ||
|
|
@@ -13,7 +18,9 @@ | |
| [](https://github.com/joennlae/halutmatmul/actions/workflows/hw_linting.yaml) | ||
| [](https://github.com/joennlae/halutmatmul/actions/workflows/hw_dv.yaml) | ||
|
|
||
| # Paper | ||
| </div> | ||
|
|
||
| ## Paper | ||
|
|
||
| * [Stella Nera: Achieving 161 TOp/s/W with Multiplier-free DNN Acceleration based on Approximate Matrix Multiplication](https://arxiv.org/abs/2311.10207) | ||
|
|
||
|
|
@@ -22,6 +29,7 @@ | |
|
|
||
| *The recent Maddness method approximates Matrix Multiplication (MatMul) without the need for multiplication by using a hash-based version of product quantization (PQ). The hash function is a decision tree, allowing for efficient hardware implementation, as multiply-accumulate operations are replaced by decision tree passes and LUT lookups. Stella Nera is the first Maddness accelerator achieving 15x higher area efficiency (GMAC/s/mm^2) and 25x higher energy efficiency (TMAC/s/W) than direct MatMul accelerators in the same technology. In a commercial 14 nm technology and scaled to 3 nm, we achieve an energy efficiency of 161 TOp/s/[email protected] with a Top-1 accuracy on CIFAR-10 of over 92.5% using ResNet9.* | ||
|
|
||
|
|
||
| # Algorithmic - Maddness | ||
|
|
||
|  | ||
|
|
@@ -56,9 +64,9 @@ mse = np.square(C_halut - C).mean() | |
| print(mse) | ||
| ``` | ||
|
|
||
| # Hardware - OpenROAD flow results from CI | ||
| # Hardware - OpenROAD flow results from CI - NOT OPTIMIZED | ||
|
|
||
| All NanGate45 results are NOT OPTIMIZED! The results are only for reference and to show the flow works. | ||
| All completely open hardware results are NOT OPTIMIZED! The results are only for reference and to show the flow works. | ||
|
|
||
| | All Designs | NanGate45 | | ||
| | ------------- | ------------- | | ||
|
|
@@ -86,8 +94,8 @@ ACC_TYPE=INT DATA_WIDTH=8 NUM_M=8 NUM_DECODER_UNITS=4 NUM_C=16 make halut-open-s | |
| | Voltage [V] | 1.1 | | ||
| | Util [%] | 50.4 | | ||
| | TNS | 0 | | ||
| | Clock Net |  | | ||
| | Routing |  | | ||
| | Clock Net | <img src="https://raw.githubusercontent.com/joennlae/halutmatmul-openroad-reports/main/latest/nangate45/halut_matmul/reports/final_clocks.webp" alt="Clock Net" style="max-width: 100%; height: auto;"> | | ||
| | Routing | <img src="https://raw.githubusercontent.com/joennlae/halutmatmul-openroad-reports/main/latest/nangate45/halut_matmul/reports/final_routing.webp" alt="Routing" style="max-width: 100%; height: auto;"> | | ||
| | GDS | [GDS Download](https://raw.githubusercontent.com/joennlae/halutmatmul-openroad-reports/main/latest/nangate45/halut_matmul/results/6_final.gds) | | ||
|
|
||
|
|
||
|
|
@@ -101,8 +109,8 @@ ACC_TYPE=INT DATA_WIDTH=8 NUM_M=8 NUM_DECODER_UNITS=4 NUM_C=16 make halut-open-s | |
| | Voltage [V] | 1.1 | | ||
| | Util [%] | 48.7 | | ||
| | TNS | 0 | | ||
| | Clock Net |  | | ||
| | Routing |  | | ||
| | Clock Net | <img src="https://raw.githubusercontent.com/joennlae/halutmatmul-openroad-reports/main/latest/nangate45/halut_encoder_4/reports/final_clocks.webp" alt="Clock Net" style="max-width: 100%; height: auto;"> | | ||
| | Routing | <img src="https://raw.githubusercontent.com/joennlae/halutmatmul-openroad-reports/main/latest/nangate45/halut_encoder_4/reports/final_routing.webp" alt="Routing" style="max-width: 100%; height: auto;"> | | ||
| | GDS | [GDS Download](https://raw.githubusercontent.com/joennlae/halutmatmul-openroad-reports/main/latest/nangate45/halut_encoder_4/results/6_final.gds) | | ||
|
|
||
|
|
||
|
|
@@ -116,8 +124,8 @@ ACC_TYPE=INT DATA_WIDTH=8 NUM_M=8 NUM_DECODER_UNITS=4 NUM_C=16 make halut-open-s | |
| | Voltage [V] | 1.1 | | ||
| | Util [%] | 52.1 | | ||
| | TNS | 0 | | ||
| | Clock Net |  | | ||
| | Routing |  | | ||
| | Clock Net | <img src="https://raw.githubusercontent.com/joennlae/halutmatmul-openroad-reports/main/latest/nangate45/halut_decoder/reports/final_clocks.webp" alt="Clock Net" style="max-width: 100%; height: auto;"> | | ||
| | Routing | <img src="https://raw.githubusercontent.com/joennlae/halutmatmul-openroad-reports/main/latest/nangate45/halut_decoder/reports/final_routing.webp" alt="Routing" style="max-width: 100%; height: auto;"> | | ||
| | GDS | [GDS Download](https://raw.githubusercontent.com/joennlae/halutmatmul-openroad-reports/main/latest/nangate45/halut_decoder/results/6_final.gds) | | ||
|
|
||
|
|
||
|
|
@@ -133,11 +141,7 @@ conda activate halutmatmul | |
| conda env create -f environment_gpu.yml --prefix /scratch/janniss/conda/halutmatmul_gpu | ||
| ``` | ||
|
|
||
| ## References | ||
| ### References | ||
|
|
||
| * [arXiv](https://arxiv.org/abs/2106.10860) Maddness paper | ||
| * Based on [MADDness/Bolt](https://github.com/dblalock/bolt). | ||
|
|
||
| ## Hackernews mention (comments only) and discussion | ||
|
|
||
| * [HN: Bolt: Faster matrix and vector operations that run on compressed data](https://news.ycombinator.com/item?id=31792206) | ||