Awesome

Solutions to Advent of Code puzzles using the Lamellar Runtime.

For most these problems, there appears to exist an efficient serial implementation, limiting the benefit of a parallel (and distributed) runtime. Still, this is an effort to explore Lamellar and provide some examples on how to use the runtime!

Generally, I will implement a serial solution as well as at least one Lamellar based solution.

I utilize the cargo-aoc crate as the execution harness.

install using cargo install cargo-aoc

run using cargo aoc optionally specifying a day with -d <day>

bench using cargo aoc bench optionally specifying a day with -d <day> and including the input parsing with -g

Performance

Run in WSL on an AMD 7950x

<table> <tr><th>Day</th><th>Part 1 Performance</th><th>Part 2 Performance</th></tr> <tr><td>

D	Implementation
1	Serial
1	Active Message


2	Serial
2	Active Message


3	Serial
3	Active Message


4	Serial
4	Active Message


5	Serial
5	Active Message


6	Serial
6	Active Message


7	Serial
7	Active Message


8	Serial
8	Lamellar Array


9	Serial
9	Active Message


10	Serial
10	Active Message


11	Serial
11	Active Message


12	Serial
12	Active Message


13	Serial
13	Active Message


14	Serial
14	Active Message


15	Serial
15	Active Message


16	Serial
16	Active Message


17	Serial
17	Active Message

</td><td>

Generator	Runner	Total
14.032 ns	16.779 µs	16.793 µs
63.054 µs	5.0559 µs	68.109 µs


13.841 ns	32.918 µs	32.931 µs
16.551 µs	10.880 µs	27.431 µs


5.4154 µs	29.827 µs	35.242 µs
24.537 µs	12.151 µs	36.688 µs


14.200 ns	270.32 µs	270.334 µs
22.012 µs	62.169 µs	84.181 µs


13.757 ns	17.769 µs	17.782 µs
13.837 ns	43.560 µs	43.573 µs


195.02 ns	14.865 ns	209.885 ns
199.93 ns	2.7972 µs	2.996 µs


13.784 ns	127.97 µs	127.98 µs
28.847 µs	159.19 µs	188.037 µs


8.8126 µs	22.189 µs	31.0 µs
42.615 µs	23.770 ms	23.812 ms


81.735 µs	23.400 µs	105.135 µs
80.600 µs	9.2678 µs	89.868 µs


14.266 ns	71.391 µs	71.40 µs
14.031 ns	161.86 µs	161.87 µs


13.945 ns	50.208 µs
13.986 ns	27.313 µs


13.806 ns	332.89 µs	332.90 µs
13.950 ns	190.53 µs	190.54 µs


47.878 µs	22.416 µs	70.294 µs
111.10 µs	10.529 µs	121.629 µs


13.949 ns	8.9763 µs	8.9893 µs
14.015 ns	26.257 µs	26.271 µs


13.783 ns	22.018 µs	22.031 µs
110.73 µs	4.0274 µs	114.75 µs


22.668 µs	466.67 µs	489.34 µs
23.969 µs	676.32 µs	700.29 µs


21.387 µs	95.200 ms	95.221 ms
20.865 µs	317.46 ms	317.48 ms

</td><td>

Generator	Runner	Total
14.011 ns	283.80 µs	283.81 µs
72.646 µs	60.733 µs	133.37 µs


13.786 ns	33.285 µs	33.298 µs
16.181 µs	9.4443 µs	27.625 µs


5.1958 µs	37.210 µs	42.405 µs
24.124 µs	11.453 µs	35.177 µs


13.715 ns	299.45 µs	313.165 µs
21.927 µs	78.105 µs	100.032 µs


13.779 ns	38.057 µs	38.070 µs
14.019 ns	54.063 µs	54.077 µs


145.28 ns	4.8219 ns	150.1 ns
146.73 ns	789.15 ns	935.8 ns


13.800 ns	148.86 µs	148.87 µs
29.071 µs	165.84 µs	183.91 µs


9.1250 µs	154.30 µs	163.42 µs
69.443 µs	66.549 ms	66.62 ms


80.005 µs	30.690 µs	110.690 µs
80.010 µs	56.556 µs	136.566 µs


14.084 ns	225.58 µs	225.59 µs
14.043 ns	180.32 µs	180.33 µs


14.030 ns	50.487 µs
14.020 ns	27.679 µs


14.052 ns	20.596 ms	20.596 ms
13.889 ns	5.561 ms	5.561 ms


49.435 µs	134.58 µs	183.493 µs
112.14 µs	51.599 µs	163.739 µs


14.087 ns	3.4685 ms	3.4685 ms
14.069 ns	6.5674 ms	6.5674 ms


13.834 ns	113.54 µs	113.55 µs
149.66 µs	34.455 µs	184.02 µs


25.022 µs	129.08 ms	129.1 ms
22.919 µs	26.667 ms	26.689 ms


21.871 µs µs	336.22 ms	336.24 ms
20.801 µs	1.1098 s	1.1098 s

</td></tr> </table>