Skip to content

Instantly share code, notes, and snippets.

@mludvig
Last active December 27, 2021 05:43
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save mludvig/31bf845061aa9d56d5de04707518fe6c to your computer and use it in GitHub Desktop.
Save mludvig/31bf845061aa9d56d5de04707518fe6c to your computer and use it in GitHub Desktop.
Instance Type GPU Type CPUs Measured MH/s Spot Price / hr MH/s per $1
g5.xlarge 1x Tesla A10g 4 56 $0.30 186
g4dn.xlarge 1x Tesla T4 4 25 $0.16 158
p4d.24xlarge 8x Tesla A100 96 1404 $9.83 143
g4dn.2xlarge 1x Tesla T4 4 25 $0.23 111
p3.8xlarge 4x Tesla V100 32 372 $3.67 101
p3.16xlarge 8x Tesla V100 64 744 $7.34 101
p3.2xlarge 1x Tesla V100 8 93 $0.92 101
g5.24xlarge 4x Tesla A10g 96 226 $2.44 93
g5.48xlarge 8x Tesla A10g 192 452 $4.89 93
g4dn.12xlarge 4x Tesla T4 48 99 $1.17 85
g4dn.metal 8x Tesla T4 96 204 $2.55 80
p3dn.24xlarge 8x Tesla V100 96 740 $9.36 79
@andremxmx
Copy link

andremxmx commented Dec 26, 2021

g5.xlarge --> $0.5082

spt price so the best cost efective rn is g5g or g4dn, thanks for your incredible help sir]!

@mludvig
Copy link
Author

mludvig commented Dec 26, 2021

@andremxmx Spot prices fluctuate. The AWS Ether miner CFN template pulls the current prices at the time of deployment and selects the most cost effective instance type for you.

@andremxmx
Copy link

Hoo nice nice thanks for dat! i just try the v520 only get 2.7 mh on eth whaiting for your post ;)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment