Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Fix: Update README.md #4

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Fix: Update README.md #4

wants to merge 1 commit into from

Conversation

coderbirju
Copy link
Owner

Issue #, if available:

Description of changes:

Testing done:

  • I've reviewed the guidance in CONTRIBUTING.md

License Acceptance

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.

@coderbirju
Copy link
Owner Author

Here's a tabular comparison of the performance difference between Finch and Docker, in terms of percentages, based on the provided benchmark data:

alpine image

Benchmark Metric Finch Docker % Difference (Finch vs Docker)
BenchmarkContainerRun ns/op 509,099,371 724,520,426 42.32% (Docker Slower)
%cpu_avg/op 58.32 37.93 (34.96%) (Finch More CPU Usage)
cpu_seconds/op 0.5090 0.7243 42.32% (Docker Slower)
disk_bytes/op 222,822 82,739 (62.86%) (Finch Higher)
B/op 464,350 649,304 39.83% (Docker Higher)
allocs/op 1,302 1,801 38.29% (Docker Higher)
BenchmarkContainerPull ns/op 520,815,319 599,878,204 15.18% (Docker Slower)
%cpu_avg/op 44.33 15.27 (65.56%) (Finch More CPU Usage)
cpu_seconds/op 0.5208 0.5996 15.18% (Docker Slower)
disk_bytes/op 10,067,968 -3,276 N/A (Docker Less Disk Usage)
B/op 469,862 550,704 17.21% (Docker Higher)
allocs/op 1,316 1,535 16.65% (Docker Higher)
BenchmarkImageBuild ns/op 909,495,694 434,906,720 (52.17%) (Finch Slower)
%cpu_avg/op 58.27 43.62 (25.15%) (Finch More CPU Usage)
cpu_seconds/op 0.9092 0.4347 (52.17%) (Finch Slower)
disk_bytes/op 10,172,825 81,920 N/A (Finch More Disk Usage)
B/op 814,846 400,912 (50.81%) (Finch Higher)
allocs/op 2,248 1,129 (49.78%) (Finch Higher)
BenchmarkImageDelete ns/op 153,457,410 82,041,749 (46.55%) (Finch Slower)
%cpu_avg/op 59.21 53.17 (10.20%) (Finch More CPU Usage)
cpu_seconds/op 0.1534 0.0820 (46.55%) (Finch Slower)
disk_bytes/op -1,984,921 819.0 N/A (Finch Less Disk Usage)
B/op 137,291 77,808 (43.34%) (Finch Higher)
allocs/op 416 255 (38.70%) (Finch Higher)

Insights:

  • Finch vs Docker: Docker is slower in execution time (ns/op) in most benchmarks, especially in BenchmarkContainerRun, where Docker is 42.32% slower.
  • CPU Usage: Finch consistently uses more CPU (%cpu_avg/op), with a 65.56% higher CPU utilization in the BenchmarkContainerPull.
  • Memory and Allocations: Docker generally uses more memory and has more allocations per operation, with 39.83% higher memory usage in the BenchmarkContainerRun.

Finch demonstrates better performance in terms of speed and CPU usage in many cases, while Docker tends to consume more memory and allocations.

@coderbirju
Copy link
Owner Author

coderbirju commented Sep 24, 2024

Here's a tabular comparison of the performance difference between Finch and Docker based on the provided benchmark data, in terms of percentages:

public.ecr.aws/y0o4y9o3/anaconda-pkg-build:latest

Benchmark Metric Finch Docker % Difference (Finch vs Docker)
BenchmarkContainerRun ns/op 587,066,481 681,067,276 15.51% (Docker Slower)
%cpu_avg/op 70.52 31.08 (55.92%) (Finch More CPU Usage)
cpu_seconds/op 0.5870 0.6810 15.51% (Docker Slower)
disk_bytes/op 196,608 -32,768 N/A (Finch Higher)
B/op 531,952 618,248 16.22% (Docker Higher)
allocs/op 1,485 1,718 15.71% (Docker Higher)
BenchmarkContainerPull ns/op 322,638,206 647,970,164 50.21% (Docker Slower)
%cpu_avg/op 37.93 11.20 (70.46%) (Finch More CPU Usage)
cpu_seconds/op 0.3226 0.6479 50.21% (Docker Slower)
disk_bytes/op 0 212,992 N/A (Docker More Disk Usage)
B/op 290,824 600,368 51.56% (Docker Higher)
allocs/op 832 1,680 50.48% (Docker Higher)
BenchmarkImageBuild ns/op 38,061,848,792 36,747,285,615 (3.44%) (Finch Slower)
%cpu_avg/op 81.51 67.71 (16.92%) (Finch More CPU Usage)
cpu_seconds/op 38.06 36.75 (3.44%) (Finch Slower)
disk_bytes/op -94,490,624 1,715,875,840 N/A (Docker More Disk Usage)
B/op 34,279,360 33,705,112 (1.68%) (Finch Higher)
allocs/op 92,556 91,012 (1.67%) (Finch Higher)
BenchmarkImageDelete ns/op 135,232,934 119,696,653 (12.91%) (Finch Slower)
%cpu_avg/op 60.61 68.33 (11.30%) (Docker More CPU Usage)
cpu_seconds/op 0.1352 0.1196 (12.91%) (Finch Slower)
disk_bytes/op -49,152 4,096 N/A (Finch Less Disk Usage)
B/op 117,592 108,608 (7.63%) (Finch Higher)
allocs/op 362 339 (6.35%) (Finch Higher)

Insights:

  • Execution Time (ns/op): Docker is slower in most benchmarks, particularly in BenchmarkContainerPull, where Docker is 50.21% slower than Finch.
  • CPU Usage (%cpu_avg/op): Finch consistently uses more CPU, with 70.46% higher CPU usage in the BenchmarkContainerPull.
  • Memory and Allocations: Docker uses more memory and has more allocations per operation, particularly in BenchmarkContainerPull, where Docker consumes 50.48% more allocations.

Finch outperforms Docker in terms of execution time and CPU usage, while Docker tends to consume more memory and resources in some cases.

@coderbirju
Copy link
Owner Author

Here’s the tabular comparison of the performance difference between Finch and Docker based on the provided benchmark data, in terms of percentages:

public.ecr.aws/soci-workshop-examples/node:latest

Benchmark Metric Finch Docker % Difference (Finch vs Docker)
BenchmarkContainerRun ns/op 3,455,454,585 7,757,108,049 (55.45%) (Docker Slower)
%cpu_avg/op 28.46 18.89 (50.65%) (Finch More CPU Usage)
cpu_seconds/op 3.455 7.757 (55.45%) (Docker Slower)
disk_bytes/op -26,255,360 -394,506,240 93.34% (Finch Less Disk Usage)
B/op 3,131,360 7,074,288 55.74% (Docker Higher)
allocs/op 8,510 19,173 55.61% (Docker Higher)
BenchmarkContainerPull ns/op 22,118,403,240 44,462,041,302 (50.24%) (Docker Slower)
%cpu_avg/op 66.26 35.29 (87.77%) (Finch More CPU Usage)
cpu_seconds/op 22.12 44.46 (50.24%) (Docker Slower)
disk_bytes/op 2,334,605,312 1,388,736,512 40.46% (Finch More Disk Usage)
B/op 19,829,840 40,553,272 51.10% (Docker Higher)
allocs/op 53,605 109,557 51.07% (Docker Higher)
BenchmarkImageBuild ns/op 76,127,741,749 24,342,207,115 (68.04%) (Finch Slower)
%cpu_avg/op 76.52 58.95 (29.83%) (Finch More CPU Usage)
cpu_seconds/op 76.13 24.34 (68.04%) (Finch Slower)
disk_bytes/op 1,498,722,304 1,293,246,464 15.89% (Finch More Disk Usage)
B/op 68,551,648 22,239,072 (67.55%) (Finch Higher)
allocs/op 185,225 60,135 (67.53%) (Finch Higher)
BenchmarkImageDelete ns/op 1,041,080,136 117,616,361 (88.01%) (Finch Slower)
%cpu_avg/op 62.38 57.67 (8.16%) (Finch More CPU Usage)
cpu_seconds/op 1.041 0.1176 (88.01%) (Finch Slower)
disk_bytes/op -1,382,612,992 -27,705,344 (98.00%) (Finch Less Disk Usage)
B/op 954,576 108,768 (88.61%) (Finch Higher)
allocs/op 2,627 339 (87.09%) (Finch Higher)

Insights:

  • Execution Time (ns/op): Docker is significantly slower than Finch in most benchmarks, particularly in BenchmarkContainerRun and BenchmarkContainerPull, where Docker is 55.45% and 50.24% slower, respectively.
  • CPU Usage (%cpu_avg/op): Finch consistently uses more CPU, with a notable 87.77% higher CPU usage in BenchmarkContainerPull.
  • Memory and Allocations: Docker uses more memory and allocations per operation in general. However, Finch has much higher allocations in the BenchmarkImageBuild and BenchmarkImageDelete, with 67.53% and 87.09% more allocations respectively.

Finch performs better in terms of execution time, but consumes more CPU in certain benchmarks, while Docker generally consumes more memory and resources, particularly in BenchmarkContainerPull.

@coderbirju
Copy link
Owner Author

Here’s the tabular comparison of the performance difference between Finch and Docker based on the provided benchmark data, in terms of percentages:

public.ecr.aws/soci-workshop-examples/tensorflow_gpu:latest

Benchmark Metric Finch Docker % Difference (Finch vs Docker)
BenchmarkContainerRun ns/op 3,128,082,038 7,451,326,779 (58.04%) (Docker Slower)
%cpu_avg/op 24.53 15.13 (62.09%) (Finch More CPU Usage)
cpu_seconds/op 3.128 7.451 (58.04%) (Docker Slower)
disk_bytes/op 241,664 -374,542,336 (64.54%) (Finch More Disk Usage)
B/op 2,862,224 6,833,928 (58.12%) (Docker Higher)
allocs/op 7,779 18,520 (58.01%) (Docker Higher)
BenchmarkContainerPull ns/op 10,246,951,854 19,823,840,965 (48.34%) (Docker Slower)
%cpu_avg/op 60.68 68.15 (10.95%) (Docker More CPU Usage)
cpu_seconds/op 10.25 19.82 (48.34%) (Docker Slower)
disk_bytes/op 1,332,887,552 1,406,275,584 (5.21%) (Docker Higher Disk Usage)
B/op 9,208,952 18,152,184 (49.28%) (Docker Higher)
allocs/op 24,920 49,073 (49.22%) (Docker Higher)
BenchmarkImageBuild ns/op 339,181,411,461 134,969,791,209 (60.30%) (Finch Slower)
%cpu_avg/op 71.97 65.45 (9.97%) (Finch More CPU Usage)
cpu_seconds/op 339.2 135.0 (60.30%) (Finch Slower)
disk_bytes/op 9,682,067,456 6,976,700,416 (38.61%) (Finch Higher Disk Usage)
B/op 307,573,008 123,845,464 (60.15%) (Finch Higher)
allocs/op 829,974 334,218 (59.74%) (Finch Higher)
BenchmarkImageDelete ns/op 1,130,480,301 132,564,220 (88.27%) (Finch Slower)
%cpu_avg/op 55.35 75.76 (26.92%) (Docker More CPU Usage)
cpu_seconds/op 1.130 0.1325 (88.27%) (Finch Slower)
disk_bytes/op -1,387,474,944 -12,288 (99.12%) (Finch Less Disk Usage)
B/op 1,031,552 118,520 (88.51%) (Finch Higher)
allocs/op 2,836 366 (87.10%) (Finch Higher)

Key Insights:

  • Execution Time (ns/op): Docker is slower in the BenchmarkContainerRun and BenchmarkContainerPull, with 58.04% and 48.34% slower times, respectively.
  • CPU Usage (%cpu_avg/op): Docker uses more CPU in BenchmarkContainerPull and BenchmarkImageDelete, while Finch uses more CPU in BenchmarkContainerRun and BenchmarkImageBuild.
  • Memory and Allocations: Finch consumes more memory and allocations in both the BenchmarkImageBuild and BenchmarkImageDelete operations, where the difference is significant—over 59% in both cases.
  • Disk Usage: Finch generally shows higher disk usage in BenchmarkImageBuild but significantly lower (more efficient) disk usage in BenchmarkContainerRun and BenchmarkImageDelete.

Finch tends to outperform Docker in execution time, although it uses more resources in certain areas like memory and allocations for some operations.

@coderbirju coderbirju changed the title Update README.md Fix: Update README.md Oct 25, 2024
@coderbirju coderbirju closed this Oct 25, 2024
@coderbirju coderbirju reopened this Oct 25, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant