Skip to content

Commit

Permalink
-K usage info
Browse files Browse the repository at this point in the history
  • Loading branch information
@madMAx43v3r
madMAx43v3r committed Jul 12, 2021
1 parent 51fea3f commit 974d6e5
Showing 1 changed file with 1 addition and 0 deletions.
1 change: 1 addition & 0 deletions README.md
View file
Expand Up @@ -36,6 +36,7 @@ Usage:
-c, --contract arg Pool Contract Address (62 chars)
-f, --farmerkey arg Farmer Public Key (48 bytes)
-G, --tmptoggle Alternate tmpdir/tmpdir2 (default = false)
-K, --rmulti2 arg Thread multiplier for P2 (default = 1)

This comment has been minimized.

Sign in to view

Copy link
@ascandal

ascandal Jul 14, 2021 •
edited

I'm not certain how to use this new -K option, especially in relation to the -r, -u, and -v options. Could we add a bit more of a option description somewhere in the readme?

This comment has been minimized.

Sign in to view

Copy link
@itestings

itestings Jul 14, 2021

As i understand by default -K is 1 so if need to double i think we should use -K 2.

This comment has been minimized.

Sign in to view

Copy link
@XeonBloomfield

XeonBloomfield Jul 14, 2021

If you set -r to 16 and -K to 2, then during Phase 2 32 threads will be running.

This comment has been minimized.

Sign in to view

Copy link
@JustMeROMs

JustMeROMs Jul 14, 2021

hmm i thought it was either 0 or 1?

This comment has been minimized.

Sign in to view

Copy link
@itestings

itestings Jul 14, 2021

If you set -r to 16 and -K to 2, then during Phase 2 32 threads will be running.

Thank you for clear reply :) now i got how this work !

This comment has been minimized.

Sign in to view

Copy link
@itestings

itestings Jul 14, 2021

hmm i thought it was either 0 or 1?

so if you have -r 16 and -K will be 0 than what we get ? so 1 and 2 is correct :)

This comment has been minimized.

Sign in to view

Copy link
@JustMeROMs

JustMeROMs Jul 14, 2021

thank you.

This comment has been minimized.

Sign in to view

Copy link
@Dardanet

Dardanet Jul 16, 2021

What I am not sure about is, do we have to take cores or threads? I tried a lot of configs. Cores, buckets and so far..

If I am using all threads, I am realitvely slow. If I take the half and add 10-20 % I have the best results..

This comment has been minimized.

Sign in to view

Copy link
@itestings

itestings Jul 16, 2021

What I am not sure about is, do we have to take cores or threads? I tried a lot of configs. Cores, buckets and so far..

If I am using all threads, I am realitvely slow. If I take the half and add 10-20 % I have the best results..

i also didn't understand how it really works but i get best results for example on SAS SSD+24threads on Xeon x6*2cpu when i use -u 7 and -n 5, mean 1 plot i get always almost same time but if i make 5 or 7-8 in a row i get them done by 15 min faster each ... so on 5 plots i win 75 min and also i get this results if i don't use -w . tried a lot of tests but cannot get too much performance on my old hardware ...

This comment has been minimized.

Sign in to view

Copy link
@w4d3j

w4d3j Jul 21, 2021

What I am not sure about is, do we have to take cores or threads? I tried a lot of configs. Cores, buckets and so far..
If I am using all threads, I am realitvely slow. If I take the half and add 10-20 % I have the best results..

i also didn't understand how it really works but i get best results for example on SAS SSD+24threads on Xeon x6*2cpu when i use -u 7 and -n 5, mean 1 plot i get always almost same time but if i make 5 or 7-8 in a row i get them done by 15 min faster each ... so on 5 plots i win 75 min and also i get this results if i don't use -w . tried a lot of tests but cannot get too much performance on my old hardware ...

The -u (and -v) flags should be powers of 2 (2^X). The default is 128, and the value just above and below are 256 and 64, respectively. I'm not sure what using 7 gets you, but give these values a try and see if it helps.

This comment has been minimized.

Sign in to view

Copy link
@gitrid

gitrid Jul 21, 2021

What I am not sure about is, do we have to take cores or threads? I tried a lot of configs. Cores, buckets and so far..
If I am using all threads, I am realitvely slow. If I take the half and add 10-20 % I have the best results..

i also didn't understand how it really works but i get best results for example on SAS SSD+24threads on Xeon x6*2cpu when i use -u 7 and -n 5, mean 1 plot i get always almost same time but if i make 5 or 7-8 in a row i get them done by 15 min faster each ... so on 5 plots i win 75 min and also i get this results if i don't use -w . tried a lot of tests but cannot get too much performance on my old hardware ...

The -u (and -v) flags should be powers of 2 (2^X). The default is 128, and the value just above and below are 256 and 64, respectively. I'm not sure what using 7 gets you, but give these values a try and see if it helps.

Only in initial releases, the default was 128, for a longer time it is 256.

This comment has been minimized.

Sign in to view

Copy link
@w4d3j

w4d3j Jul 23, 2021 •
edited

What I am not sure about is, do we have to take cores or threads? I tried a lot of configs. Cores, buckets and so far..
If I am using all threads, I am realitvely slow. If I take the half and add 10-20 % I have the best results..

i also didn't understand how it really works but i get best results for example on SAS SSD+24threads on Xeon x6*2cpu when i use -u 7 and -n 5, mean 1 plot i get always almost same time but if i make 5 or 7-8 in a row i get them done by 15 min faster each ... so on 5 plots i win 75 min and also i get this results if i don't use -w . tried a lot of tests but cannot get too much performance on my old hardware ...

The -u (and -v) flags should be powers of 2 (2^X). The default is 128, and the value just above and below are 256 and 64, respectively. I'm not sure what using 7 gets you, but give these values a try and see if it helps.

Only in initial releases, the default was 128, for a longer time it is 256.

Thanks. To clarify, I meant the official Chia plotter defaults to 128. But regardless of those, 7 is not the value you should probably be using.
--help Print help
```

Expand Down

2 comments on commit 974d6e5

@romeo88vn
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

can you make another version or integrate it into the next version for plot size k31?
Thanks!

@altendky
Copy link

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

can you make another version or integrate it into the next version for plot size k31?
Thanks!

One of various duplicates of #892.

Please sign in to comment.